Patent Application
20250007643
To transmit information efficiently in a communication system, the available resource may be cautiously used with a good spread of energy over the allocated bandwidth and duration. Ideally, for efficient low-cost low-power transceivers, packets should have a reasonable bandwidth and limited power variation over time, i.e., low Peak-to-Average-Power-Ratio (PAPR), and provide solutions to be easily modulated and detected/synchronized/demodulated, even at low Signal-to-Interference-plus-Noise-Ratio (SINR) for a wide range of scenarios. Many current systems are not so scalable, implemented with tradeoffs or worst-case design settings that can increase complexity and power consumption, and waste resources; this reduces the global wireless network performance in real environments.
In one aspect, an apparatus includes: a circuit to generate a packet having a preamble, a header, and a data portion, where at least the preamble and the header are dynamically adaptable (and adapted) based at least in part on a range between the apparatus and a receiver; and a radio frequency (RF) front end circuit coupled to the circuit to process and transmit the packet.
In an implementation, the apparatus further comprises a controller coupled to the circuit, where the controller is to determine a length of the preamble and a length of the header based at least in part on the range. The circuit may generate the preamble having: a first field with a first variable length, the first variable length based at least in part on the range; and a second field to identify the packet and enable validation of detection of the packet by the receiver. The circuit may generate the first field having a sequence that is repeated a plurality of times to form the first variable length.
In an implementation, the circuit may generate the header comprising: a first field with a second variable length, the first field to identify a range mode based on the range and to define a structure of the packet; and a second field with a third variable length. One or more of the fields may be used to progressively scale the range and group information over the packet, optionally refined by each field. In an example, the second field is to refine the range mode to identify a modulation and coding scheme (MCS) selection for a third field of the header, and to provide the group information. The circuit may modulate the second field of the header and not modulate the first field of the header. The controller may determine the range and the group information based at least in part on prior exchanges between the apparatus and the receiver, to adapt one or more link parameters.
In an implementation, the circuit may generate the header comprising the third field with a fourth variable length, the third field to identify a MCS selection for a data portion of the packet and to indicate a length of the data portion of the packet. The circuit may modulate the third field of the header according to the MCS selection for the third field obtained from the second field of the header. The circuit may form the data portion of the packet comprising a symbol, the circuit to: modulate the symbol with a sequence; extend the modulated symbol to obtain a plurality of modulated symbols; and perform, on the plurality of modulated symbols, a plurality of operations according to a Recipe of operations, to obtain extended and modulated symbols for the RF front end circuit to transmit to the receiver.
In an implementation, the controller may determine the sequence and the Recipe of operations based on one or more of the range, the MCS selection, or the group information. The controller may dynamically adapt at least one of the range, the MCS selection, or the group information for another packet to be transmitted from the apparatus to the receiver.
In another aspect, a method comprises: receiving, in a receiver, a RF signal comprising a packet and processing the RF signal to obtain a digital stream comprising the packet; identifying, from the digital stream, a range mode based on range information included in a first field of a header of the packet, the first field having a first variable length based at least in part on the range information; identifying, from the digital stream, a group based on group information included in a second field of the header of the packet, the second field having a second variable length based at least in part on the range information; identifying a length of a data portion of the packet based on length information included in a third field of the header of the packet, the third field having a third variable length based at least in part on the range information; and demodulating the packet according to at least one MCS, the at least one MCS determined based at least in part on the range information.
In an implementation, the method further comprises progressively refining MCS information using one or more of the second field of the header and the third field of the header. The method further comprises demodulating at least one of: the data portion of the packet according to the progressively refined MCS information; and the second field of the header according to the progressively refined MCS information. The method may further comprise: correlating a preamble of the packet to detect the preamble, the preamble comprising a first field having a variable length formed of a sequence that is repeated a plurality of times, the variable length based at least in part on the range information; and validating the detection of the preamble of the packet using a second field of the preamble. The method may further comprise dropping the packet and not demodulating at least the data portion of the packet based on demodulating one or more of the first field, the second field, or the third field of the header.
In another aspect, a system includes a transceiver comprising: a circuit to generate a packet having a preamble, a header, and a data portion, where the preamble and the header have scalable lengths based at least in part on a range between the system and a receiver, and: the preamble comprises a first field with a first variable length based at least in part on the range, and a second field to identify the packet and enable validation of detection of the packet by the receiver; and the header comprises a first field with a second variable length, the first field to identify a range mode based on the range, a second field with a third variable length, the second field comprising group information, and a third field with a fourth variable length, the third field to indicate a length of a data portion of the packet. The transceiver may further include an encoder coupled to the circuit, the encoder to encode the packet comprising the header and the data portion into a plurality of symbols; and a modulator coupled to the encoder, the modulator to: modulate the plurality of symbols using at least one MCS, the at least one MCS based at least in part on the range; and a RF front end circuit coupled to the modulator, the RF front end circuit to form a RF signal comprising the modulated plurality of symbols and transmit the RF signal comprising the modulated plurality of symbols.
In an implementation, the circuit is to generate the header comprising at least some of the MCS information, the at least some of the MCS information comprising progressively refined MCS information. The modulator may be configured to modulate the header according to a first MCS based on the progressively refined MCS information and to modulate the data portion according to a second MCS based on the progressively refined MCS information.
In various embodiments, a transmitter is configured to perform modulation of communications for long-range scenarios. Although embodiments are not limited in this regard, as an example a transmitter can be configured to dynamically control modulation aspects to enable transmission of wireless signals over relatively long ranges, e.g., of up to approximately a few tens or hundreds of kilometers.
In embodiments, Constant Amplitude, and in particular Constant Amplitude Zero Auto-Correlation (CAZAC), sequences may be used to perform modulation. In general, a CAZAC sequence is a polyphase sequence having periodic complex-valued signals with a modulus of one and out-of-phase periodic (cyclic) auto-correlations equal to zero. In an embodiment, an example for construction of such signal sequences may be a combination of well-chosen quasi-CAZAC sequences and Orthogonal Frequency Division Multiplexing (OFDM) principles. The chosen CAZAC sequences may be close to Zadoff-Chu (ZC) sequences, or more generally Generalized Chirp Sequences (GCS). Such a GCS-precoded or ZC-precoded OFDM signal can be used as a basis to define a modulation scheme robust for long-range scenarios. These sequences may optionally be transformed into another domain using Fourier Transforms (Discrete Fourier Transform (DFT) or Fast Fourier Transform (FFT) processing) of OFDM systems, with potential further signal processing to shape the signal (e.g., windowing, filtering, power compression), to provide desired properties. Such sequences are referred to herein as Sweep-Sequences (SS), or more generally “sequences” and may be formed of tens to thousands of complex-valued samples.
A Generalized Chirp sequence (GCS) of complex samples, of length Nzc and parameters (P, Q, U, S), could be defined with the following equation:
where P, Q, U, S do not need to be integers.
As a subset, a root Zadoff-Chu sequence of complex samples, of length Nzc and root u, with potential cyclic shift q is defined with the following equation:
where c is generally 0 for even Nzc and 1 for default odd Nzc (ensures cyclicity/continuity if repeated); u is an integer identifying the ZC sequence, called a root index. ZC is a particular case of GCS with S=0, P=0, integer u=-U, q =-Q/2 when Nzc is even, and q=−(Q/2+U/2) when Nzc is odd.
These ZC are polyphase sequences with perfect periodic auto-correlation function (i.e., complex sequences with unit amplitude and null auto-correlation for all cyclic shifts except zero). However, compromises considering pool of sequences with low shifted auto-correlation and low inter-correlations could be sufficient for a given application.
OFDM (I) DFT conversion for Nf complex samples is in accordance with the following equation:
DFT (or FFT) processing ensures signal continuity at sequence edges, and also allows resampling (final SS bandwidth can differ from initial/sampling bandwidth, filling edge FFT bins with zeros, e.g., when Nf>Nzc).
ZC-OFDM is the combination of these two processing functions, with potential intermediate-and/or post-processing to shape the signal (e.g., centering, filtering, windowing, shifting, compressing in power). After the selection of sizes (Nzc, Nfft=Nf.L) and sampling rates (that also define the SS bandwidth and duration), there can be several imaginable sequences that are quasi-orthogonal (up to Nzc for ZC-OFDM with integer u). Note that sequences can be L-oversampled to further improve the signal shaping and properties (e.g., finer tuning/filtering and management of PAPR/Out of Band (OoB) using higher FFT of length Nfft=NfL instead of Nfft=Nf≥Nzc).
As another subset of GCS with U=0 and S=0, ultra-narrow-band simple carriers (constant frequency per block) sequences or hopping single-carrier sequences are also particular cases with low PAPR and some capabilities to provide several orthogonal sequences (or semi-orthogonal once different coding/hopping and/or Recipe of operations are applied for each user), e.g., for different values of even Q in [0, . . . Nzc] in the GCS equation above. Piecewise-non-sweeping ultra-narrow-band sequences may appear less robust, for a given SS, than the presented CAZAC sequences regarding some noises and interferences, but robustness and good use of the allocated spectrum can be recovered from the Recipes described later in this document. Then, hopping subcarriers (piecewise pure sinusoids) to design a list of sequences is also a low-rate solution to share a scarce resource, and is equivalent to precoded-OFDM with the precoding vector of size Not being all zeros except one element at once (1 complex value on the unit circle). However, a general receiver for that case may need different algorithms with potential higher complexity to demodulate or synchronize, just as less optimal spectrum (and potential dependency on filtering processes) compared to ZC and ZC-OFDM sequences. Moreover, presenting a different spread of energy over time, depending on the considered averaging periods to compute the spectrums, solutions that are more ‘peaky in frequency’ could be more restricted in power by regulations to access some wireless channels/bands.
As another intermediate example list of sequences, a subset of GCS with low sweeping slopes (non-integer 0<|U|<1 and/or more zeros on edges in a following OFDM FFT) and S=0, can provide sweeping sequences with low PAPR and capabilities to provide several orthogonal sequences (or weakly correlated) with another balance on robustness to interferences in time/frequency, number of simultaneous packets on the same band, and management of Groups. The related spectrum per SS does not spread the energy as equally as GCS/ZC-based sequences with integer U≥1, but the description of Recipes in next processing stages will allow to further spread properly the signal over the whole system bandwidth, while still allowing simplified receiver processing of sweeping sequences in real channels (with multipath interference). Simple receivers with robustness to any noise, including localized interferers ‘peaky-in-frequency’ and ‘peaky-in-time’, are possible with weighted aggregation of such blocks from the modulation. Then, from presented equations and solutions, users in the field can readily adjust parameters and durations to meet the requirements and regulations.
In one or more embodiments, a transmitter may generate and/or store a pool of such sequences. These sequences may be particularly optimized for long range wireless communications, with particular properties including the following. The sequences may enable good shapes in time and frequency that are better than generic (worst-case) OFDM, having low power variations over time (low PAPR), and good shaping in frequency (aiming ideal spectrum for the application, with low Out-of-Band emissions), as presented above. These sequences may further have good auto-correlation and cross-correlations figures, i.e., localized auto-correlation (almost null except for zero-shift auto-correlation).
Collectively, the list of sequences can be rather orthogonal (i.e., with low correlations with any shift among the list) to efficiently identify a given SS and its shift. Further optimizations with the selected modulation may include continuity when repeated (ZC, ZC-OFDM, and any precoded-OFDM sequence are already circular/continuous sequences); low impact when modulated (reduced edge effects like discontinuities involving out-of-band emissions, modified PAPR and lost power); and simple demodulation solutions (either in frequency or time domain, generally based on frequency-sweep-like properties, and/or symmetries).
Note that ZC values are naturally quasi-orthogonal when considering different integer root values u, and a couple of sequences with values of u generally corresponding to low sweeps (e.g., {1,2,3}) per block may offer good trade-offs with rather low PAPR once frequency-transformed. Note that the DFT of a ZC sequence, with Nzc prime, is still a ZC sequence (keep constant amplitude with good correlation properties), while there is no intermediate processing.
Several properties like symmetries/repetitions can exist among sequences (e.g., the conjugate of a sequence is generally its reversed-ordered version when sequences are anti-symmetric, generally representing a frequency sweep in the other direction and u′=Nzc−u) and could also save memory. These sequences are often close to frequency-sweep sequences (CSS could be seen as a particular case of ZC-OFDM with the simplest sweep over the full Nzc=Net, with u=1 and without windowing), i.e., could approach one or several tones sweeping (less than one, one or several times) on the allocated area (given bandwidth and sequence duration).
Using such a SS as a basis, a Symbol Modulated and Extended (SME) can be generated to efficiently transmit information. SME is the combination of a modulation of the SS and an extension of that signal that is repeated (e.g., an integer number of modulated SS like 4 repetitions to simplify description). In addition, each of these extended SS's may have one or more operations (from Recipe) performed on it, as described further below.
Note that in one embodiment, the same SS is used for the repetitions in the given SME (even if a potentially different operation is performed on each repeated SS inside the SME). In an embodiment to simplify, a limited set of operations inside a SME may be used. For example, in one implementation, these operations may include: time reversal (and/or conjugate operator), i.e., SS′(n)=SS (N−n), n=0, . . . N−1; phase operations (e.g., extended generic case of +/−{1,j}), i.e., SS′(n)=exp (j.θ).SS (n); frequency shifts (e.g., from a numerically controlled oscillator (NCO)), i.e., SS′(n)=SS (n)· exp (j.2TT.8.n/N) with typically |δ|≤1 if the small/fractional frequency shift is mainly used to improve signal continuity at edges or spectrum use, or any larger δ if frequency shifts are also used to manage interference-mitigation; and/or a combination of any of these three operations.
Note that SME extension is presented as a repetition of an integer number of SS to simplify description, but could also be slightly extended, similarly to OFDM Cyclic Prefixes, to better manage SME edges in real channels with echoes. As an example, instead of a cyclic extension of 3 times the SS, to have 4 cyclic repetitions of the SS, there can be 4.25 cyclic repetitions of the SS (to have a quarter symbol to smooth the signal and minimize inter-symbol interferences when there are echoes and filtering).
Note that in embodiments, SS inversion (minus) and pi/2-rotation (j) are particular cases of the phase operator, and conjugation is generally a particular case of time-reversal (sequence played in the other direction, when sequences depict symmetries) with a potential frequency-shift.
An additional variation of the selected SS and/or Recipe of operations at each new SME is possible to further extend Packet Interference Mitigation (PIM) in heavy network-loaded contexts (to increase simultaneous users without high performance drop). Encoding using adapted bit-level forward error correction (FEC) and interleaving may also be used to tolerate remaining spread errors and local interferences impact. As an example, a proper block FEC spreads information bits as much as possible on the considered block of the frame. Generic block encoding/decoding may be used (e.g., for packet headers), to adequately spread (interleave/encode/map) the information bits to symbols with a priori equal ‘information weight’ to be robust in the channel. In a particular embodiment, each information bit (after BIC encoding) is spread on several symbols, and several bit-levels (from MSB to LSB, as a symbol, among 2M, can contain M bits) of these symbols, to leverage/mitigate potential biased SINR per information bit. In practice, precomputed block matrices G (size K×N) can be predefined for each potential K/N couple of headers. But a combination of smart interleaving/FEC encoding and symbol mapping to spread information bits is also possible.
The combination of SS repetitions and operations on these repetitions inside SMEs enables an increase in robustness to channel effects, noises, blockers, and especially interferences from the same kind of packets transmitted simultaneously on the same bands (e.g., of other users). At the same time, simple and efficient recombinations can be performed at a receiver having a priori knowledge of the transmitter modulation schemes and recipes. Embodiments may also present a good spread of energy over the dedicated bandwidth and SME duration.
Moreover, another optional feature considers progressive knowledge/refinement of a user group/ID along the packet (from signaling parts to final payload) allowing a receiver to efficiently receive expected packets or stop demodulation as soon as the frame is not going to succeed for the given user, improving power-savings.
Embodiments implementing Sweep Wide-Area Modulation (SWAM) can be seen as a modulation technique for an entire packet or one or more portions of a packet (e.g., a data portion). Depending on implementation, such precoded-OFDM modulations may be included as an extra mode or “modulation and coding scheme” (MCS) of a Wireless-Smart Ubiquitous Network (WiSUN-OFDM), or as a building block of a new and optimized system/standard, more flexible, and superset of a Large-Scale Low-Rate Network (like IEEE 802.15.4g WiSUN).
Referring now to
As shown, incoming information in the form of an input bit stream (which may be generated within digital circuitry of the transmitter or a device coupled to the transmitter) are input into a scrambler 110. In various embodiments, scrambler 110 (which can be optional in some embodiments) may scramble and/or whiten the incoming bits, which are then provided to a bit-to-symbol encoder (generally encoder 120). In the high-level view of
Still with reference to
A resulting series of modulated and extended symbols (SMEs), optionally shaped/filtered/smoothed, is provided to a radio frequency (RF) front end circuit 140. Various digital and analog processing may be performed, including upconversion to RF level, filtering, gain control and amplification (including in a power amplifier of RF front end circuit 140), to result in a transmit RF signal that is transmitted via antenna 150.
Still with reference to
In embodiments, controller 160 may, based on various parameters, control selection of the type of scrambling, encoding, interleaving, reordering and symbol mapping. In addition, controller 160 may, based at least in part on input parameters and/or symbol index, determine an appropriate SS to be used for modulation, and determine an appropriate Recipe of operations to be performed on the SS. Still further, controller 160 may be configured to control any optional signal shaping performed in SWAM modulator 130. Although shown at this high level in the embodiment of
Referring now to
Method 200 demonstrates an overall flow for generation and communication of a packet from a transmitter to one or more receivers. In general, the flow of method 200 is illustrated for operations that can be performed iteratively for different portions (frames) of an overall packet.
As shown, method 200 begins by receiving information parameters (block 210). These information parameters may be received in a controller of the transmitter. Such information parameters may include certain fixed information parameters for a given type of modulation scheme. In addition, these information parameters may include information received from a network, and feedback information (or measurements) regarding (prior or current) communications from/to the transmitter.
Particularly relevant information parameters may include range information (regarding an estimated signal quality or distance between the transmitter and one or more receivers), group information, and “modulation and coding scheme” (MCS) information. Next, control passes to block 215, where the controller may determine a SS and a Recipe of operations based at least in part on at least some of the information parameters and a symbol index. As will be described, this determination is thus a dynamic determination that may be performed for each extended block of information to be communicated. The determination of blocks 210 and 215 may be performed by a controller of the transmitter.
As further shown in
Still with reference to
Still with reference to
Still referring to
Instead when a packet is completed, control passes to block 280 where multiple frames of the packet are concatenated and the extended and modulated symbols of the frames of the packet, optionally shaped/filtered/smoothed, are output to an analog RF front end circuit for upconversion and then transmission via a power amplifier and antenna. Although shown at this high level in the embodiment of
In one or more embodiments, SS's can be constructed with optional features to provide sequences having desirable properties that enable flexibility for different environments (e.g., ranges, simultaneous users, packet interference mitigation, and so forth). In an embodiment, a sequence may be generated using a GCS equation as described above. Additional operations optionally can be performed. For example, a portion of a sequence may be extracted/processed, such as performing a centering/shifting, removing one or more central values (DC=mean), ensuring symmetry, filtering edges, and/or adding zeros at edges. Understand that in different use cases, one or more of these optional operations may be performed to derive a sequence. In some implementations, this sequence may further be transformed into the frequency domain via a FFT operation. Further derivations can be performed on a transformed sequence using operations like Recipes (e.g., via a numerically controlled oscillator (NCO)) and potential windowing, to provide the final SS.
When a given SS list is selected for use in modulation, SMEs are constructed. For each symbol, a given SS is selected from the SS list using Group information and optionally symbol index. Then the SS is modulated with the symbol value (e.g., via a circular time shifting and/or a frequency shift (using an NCO)). The resulting modulated SS is extended to SME size (e.g., with CPx copies). Then based on PIM information and symbol index, a set of operations can be performed according to a Recipe of operations. Finally, potential edge processing and/or additional windowing/filtering may be performed (which may depend on SS and selected Recipe).
Referring now to
While these figures provide only limited illustration regarding the Recipe of operations in the complex domain (only displays absolute power here), they provide an example of possible solutions to manage continuity, shaping optimization, orthogonality, and spread of energy in time/frequency.
Referring now to
Referring now to
As shown, controller 360 can receive a variety of parameters, including information received from a network as well as feedback information and/or measurements. In an embodiment, the network information may include framing structures, tables of possible MCSs for a given Range, Groups and Recipe information and ways to select among a subset, and the feedback information may include channel characteristics (or measures) to select the Range and MCS and keys to define the Group information. In an embodiment, the Group information may be a value for a user in a cell area be as low as possible interfering with other signals, allowing selection of adequate SS and Recipes (avoiding similar waveforms in the same area).
In one or more embodiments, the main parameters for determining a modulation configuration are derived from Range, MCS and Group parameters. In an embodiment, the Range parameter defines the SS duration (size of building blocks), main packet structure and MCS pool definitions. The greater is the range, the better is the sensitivity and longer is the packet for the same content to transmit.
In different use cases, one or more of Range, PIM information and symbol index may be used in determining a SS sequence list to be used and what extension and Recipe of operations are to be performed. For example, controller 360 may determine configuration of modulation based on range only when there is a single or small number of users (e.g., less than a first user threshold) located at a close distance with respect to transmitter 300 (e.g., less than a first distance threshold). As distance (e.g., greater than the first distance threshold) and/or number of simultaneous users increases (e.g., greater than the first user threshold), range and PIM information may be used. And as further users and/or distance increases (e.g., greater than another distance and/or user threshold), range, PIM information and symbol index may be used to determine modulation configurations.
In light of the above discussion and depending on environment, there may be flexibility for at least three dimensions/types of information parameters to adjust the modulation to the environment/application: including Range, MCS (to select at least nb bits per symbol, i.e., nb modulation e.g., circular-shift possibilities per SS to handle channel echo impact, and FEC details and rate) and Group information. In this way, embodiments enable adaptation of Range, MCS, and Group for packet interference mitigation, to target a low error probability in any environment and condition.
Referring now to Table 1, shown is an example of MCS pool definition in accordance with an embodiment, e.g., for long-range scenarios with SS sequences of size N/f*L=128*2. The number of repeated SS per SME is equivalent to x+1 in the CPx notation of this Table; e.g., “CP3” means a cyclic-prefix of 3 corresponding to 4 SS per SME. In this arrangement, here we consider TimeShift modulation with 2 to 64 shift states depending on the MCSy index. However, with more than 16 states (4bits per FSS), performance can be degraded if channel impulse response is large (interfering echoes), generally limiting high MCS to high-throughput transmissions in favorable scenarios (low CIR) to keep receivers simple. Shown in Table 1 are 8 MCSs, which may be encoded using 3 additional MCS-signaling bits per given RANGE (here is a range with sequence length 128).
Following Table 1, shown is Table 2, which is an example definition for a particular MCS index (index 2).
Still with reference to
The resulting symbols (e.g., X symbols) are provided from encoder 310 to a modulator 320. Controller 360, based at least on the MCS information, configures modulator 320 to perform a modulation of the incoming symbols using a given SS, namely a selected SS from the list of SS's (for the packet group) from a plurality of lists of SS's that are stored in a pool of sequences. In an embodiment, modulator 320 uses a selected SS (selected by controller 360 as described below) to modulate (a single) symbol. In an embodiment, this modulation is a circular shift of the selected SS.
As shown in
In embodiments, the pool of sequences includes sets of SS that are predefined and/or computed, and which have good properties, in particular good auto-correlation and inter-correlation properties. In this way a SS and its circular-shift can be identified, having localized/peaky circular auto-correlation, and low circular inter-correlation with any shift of other SSs of sequence lists 372.
Variations in sequence lists 372 may be used to enable flexibility in packet communications to accommodate different numbers of simultaneous users. For example, different sequence lists having a same number of values (but differing values) can be used to provide concurrent or simultaneous transmissions to multiple users in an environment. And longer SS may be used for communications with more distant users within a wireless environment.
Depending upon implementation, the pool of sequences may be stored in a non-volatile memory where the pool is predetermined and stored as part of firmware. In other cases, the SS may be dynamically determined and stored, e.g., in a lookup table, or even generated ‘on-the-fly’ or dynamically, using the GCS/ZC equations described above.
In an embodiment, oversampled versions of predefined SS sequences can be saved in memory 370 to further optimize and pre-shape the signal. In some embodiments, a typical twice-rate definition (either in time or frequency) could be used as a basis for next modulation steps. Note that SSs can be selected to generate signals close to constant-amplitude modulations, and it is even possible to compress the final signal amplitude without modifying modem generic algorithms and without significantly reducing performance. In other embodiments, a subset of SS sequences can be stored in memory 370 (to reduce memory consumption) and then controller 360 can use these stored subsets to generate a complete sequence. For example, memory 370 may store a portion of a given sequence, and one or more processes may be applied to this stored sequence (e.g., symmetry: half-size and reverse to obtain a whole sequence). In addition, possible operations (e.g., conjugate), and/or filtering/resampling/FFT-OFDM may be performed to obtain a final SS used for a given SME.
In an embodiment, the SS selection may be based at least in part on a sequence size, which can be determined from range information, and may further be based on PIM2 parameters from the group information. In an embodiment, the range information may be used to obtain a list of SS (with defined size Nfft), and the PIM information (from group information) along with a symbol index may be used to select the SS from this list of SS.
Still with reference to
Depending on implementation, the resulting SMEs can transmit information based on SSs using: shifts (time or frequency cyclic shifts of SSs); complex-multiplication (IQ) on the SS signal (including potential inversion, phase modulation, amplitude modulation); and different SSs over time. In some embodiments, shifts are considered for modulation only, to allow other dimensions to be used for PIM.
To simplify processing, choices can be made at early design stages. As an example, for one implementation a Sweep Scalable Simultaneous-access Wide-Area Network subset system is defined with a given modulation scheme that is based (only) on circular shifts of the SS. Amplitude modulation is not used (to keep PAPR low). Other degrees of freedom are not used to modulate information but to increase robustness through PIM techniques described below. In other words, potential further operations on the SS (including IQ processing like signal inversion, phase shift and drift, and SS variation or conjugation) can be used in combination with repetitions to leverage interferences and noises (instead of for modulating data information), allowing simple but efficient and future-proof receivers.
In turn, the group parameter extends the signature to manage PIM (defines scrambling and PIM indexes). In one embodiment, the group information may be mainly derived from previous exchanges (network association, previous packets) and/or previous portions of a packet (header containing fields that define the group selected for the next packet portion) provided within some part of a packet, and may be divided in two subsets, PIM1 and PIM2. The PIM1 parameter can be used to define an intra-SME Recipe (vector of operations for the given MCS depends on this PIM1 index). In turn, the PIM2 parameter can be used to define an inter-SME potential variation of the selected SS and/or Recipe (e.g., select the SS over the frame, such as a constant SS index or a repeated vector of SS indexes among the SS list to further scramble interferences).
The resulting modulated symbol is extended a selected number of times in an extender 332, and is provided to a Recipe operator 334 that applies a Recipe of operations on each copy of the modulated symbol. In embodiments, extender 332 may copy the modulated symbol a given number of times (e.g., based on MCS CPx, with recipe length of usually equivalent length, defined at least in part on PIM1 information). In various embodiments, Recipe operator 334 may be configured as a programmable processor, which can be implemented within or apart from modulator 320. Additional processing such as windowing, filtering or so forth may be performed in an optional edge processor 336. While in
In embodiments, controller 360 determines this Recipe of operations based at least in part on the PIM1 information, which may be based at least in part on the group information of the information parameters. The resulting extended and modulated symbols may further be signal shaped within Recipe operator 330 (where this determination may be based on additional PIM1 information).
Referring now to Table 3, shown is an example of operations performed on a SS (for SME construction).
Note that time-reversal instead of conjugation can help in phase continuity, and then decrease unwanted emissions (out-of-band spectrum regrowth) when filtering in the front-end. Although 144 operations are presented in this example, other sets of operations are possible in other embodiments, in particular extending the allowed values of delta for frequency shifts.
Referring now to Table 4, shown is an example of Recipe definitions for an example SME construction. As an example, for the 5 rows shown (which may be selected based on a PIM1 index), when the current MCS assumes 4 SS per SME (‘CP3’ like MCS {3-7} in Table 1), the values in Table 4 correspond to the listed operations to be performed on the SSs according to a given Recipe of operations.
Note that the processing and operations can be performed in the module/phase domains instead of I/Q quadrature domains (usual representation of complex values) through polar conversion; for constant amplitude targets, this could be implemented with phase modulators instead of IQ modulators. This process can simplify the phase continuity at edges.
As in OFDM, a circular time-shift of a SS is equivalent to a linear-varying phase shift in the frequency domain (with the DFT conversion of OFDM), which is, in the case of linear-frequency-sweeping SS, similar to a circular frequency-shift. A transmitter as well as a receiver may leverage this property to simplify demodulation and perform algorithms either in time or frequency.
With embodiments, robustness in real environments is improved, not only using the presented schemes (SS, PIM), but also considering sufficient minimal shifts between symbol depending on the Channel Impulse Response (CIR) (equaling echo spread), as presented in MCS tables.
Robustness and compromises in real channels (multipath and multi-user) can be increased using: cyclic-prefix-like extension of the signal (e.g., full repetitions of the signal, or at least large compared to conventional OFDM, which often considers only a small fraction of the symbol as a cyclic extension) to simultaneously handle channel spread/echoes and potential spreading groups coding over repetitions; variations of the selected SS, among the SS list, over time; and encoding and spreading at several levels. Such operations may avoid interference with similar packets, transmitted simultaneously in a real multi-user system.
A SWAM-modulated block is thus a succession of SME symbols where a SME is constructed from: a selected SS (with a given duration and sequence content); a modulation (i.e., symbol information to transmit; 1≤Nslog2 (Nfft) bits per SME); and a Recipe, i.e., the operator scheme over the SS repetitions (e.g., a vector of operations).
Embodiments may dynamically determine generation of a SWAM block to balance throughput and robustness (both in terms of range/SNR/echoes and interference capabilities). As discussed above, all of Range, MCS, PIM information and symbol index may be used. In this way, communications may target very low sensitivity, and leverage robustness, while still being efficient as it is flexible and scalable when the context permits smaller packets at higher throughputs. In one or more embodiments, SME and payload block construction process may be performed independently for each part of a frame communication (e.g., one or more headers and a data payload).
Still with reference to
Note that transmitter 300 may transmit a packet as a collection of frames, where there may be different frames for preamble, training field, header1, header2, and data (as an example). Each frame may have its own settings (defined from fixed/top-level common parameters and/or a previous frame). In turn a frame may be constructed with a given number of SMEs. Headers generally have a fixed and known number of SMEs for a given MCS. A data portion of a packet may be formed of a variable number of SMEs (each packet having different number of bits to transmit, where a data length is given in the headers).
In addition to the flexibility afforded by the above parameters, controller 360 also may enable greater reliability in difficult environments using sub-block interleaving. Thus in some implementations, an optional sample-subblock-level interleaving may be performed. Such interleaving may be performed to increase robustness in difficult channels with potential varying-over-time noises and interferences (e.g., localized/peaky or impulsive noises). This interleaving is performed on the samples of each SME over the packet, as it is often too late at bit-level to deinterleave information that is already poisoned by localized high-level noises that give high but wrong reliabilities to a given symbol at the receiver.
Thus to provide additional robustness, optional sample-subblock-level interleaving may be performed to spread ‘sample-level information’ over a packet by splitting each SME into P subblocks. In one or more embodiments, this optional feature can be activated on a per packet basis (e.g., via a bit field option in a header of each packet) or fixed by the network or from a design choice for a given application.
Referring now to
In the example shown, frame 420 has a configuration with a simple spread over 3 SMEs to generate and send information with a simple interleaving of SME1A, SME2A, SME3A, SME1B, SME2B, SME3B, SME4A, SME5A, SME6A, SME4B, SME5B, SME6B, SME7A, SME8A, SME7B, SME8B. As seen, the first operation is to group per block of L SME (where L=3 here), with the last block for a given frame length being of size between 1 and 3. And the second operation is a simple reshaping/reordering per P sub-blocks, where P=2 here, that splits each SME into P parts equally interleaved within the frame.
The optional features to split into sub-blocks in time or in frequency (e.g. interleave in time SMEs as presented on
In general, demodulation of received packets may proceed in a reverse fashion to transmission, based on information parameters as discussed above (including Range/MCS and Group). Depending on implementation, this information may be fixed (e.g., standard/network/negotiated), or based on information obtained in a previous part/frame of a packet (e.g., to refine MCS and group information progressively). Once an initial packet detection and synchronization process is performed, the information parameters may be used to perform demodulation of each SME as follows.
First, the SS and Recipe of operations as used on the transmitter side to modulate symbols may be determined based on the information parameters and symbol index. Note that there can be a potential combination of several subparts of SME. In general, the operations performed on the transmit side may be performed in reverse order, i.e., potential edge processing (e.g., windowing/removal/weighting) may be performed, then a reverse Recipe of operations is performed, and finally the resulting SSs of the SME are aggregated to form only one received “denoised” SS.
Once the SS is recovered, it is processed either in frequency (after a FFT operation) or in time. Then a correlation is performed with an expected SS to find a most probable modulation. In a frequency domain, the cyclic correlation of sequences is equivalent to a dot product of their Fourier transform (after complex conjugation or time-reversal of one sequence). In this operation, a symbol shift can be easily derived from analysis of an inverse FFT (IFFT) operation on this correlation.
To understand, in an example without cyclic-shift (modulation0), the correlation of a sequence with itself (conjugate) results in a flat sequence in phase/angle (and even is a sequence of 1 if the SS has complex values of module 1 like CAZAC sequences), related to an IFFT with a peak in index 0 (frequency shift=0). For Sweep-Sequences, the higher the modulation shift is at a transmitter, the more cycles of sinusoids are present in a correlation vector, resulting in a related FFT with a peak at an increased index that gives an initial modulation cyclic shift. This cyclic shift is used to find/estimate the most probable transmitted cyclic shift among the 2M symbols, or to derive soft values of probabilities (reliabilities) for each potential 2M symbols in a very noisy reception environment.
This process may proceed in a loop as long as there are more symbols to process. Then, the most probable transmitted bits are deduced from a soft processing (similar to soft-FEC decoding) on the previous soft values per symbol, as a reverse operation of the transmit BIC operation to provide the transmitted bitstream.
In other embodiments, a different receiver may be used to improve performance. As an example, knowing/estimating CIR, especially in case of channel echoes, principles similar to a rake receiver can be used to better recombine echoes with adequately corrected correlations on several main paths. Note that receiver algorithm simplifications may occur in case of linear-frequency-sweep sequences.
Referring now to
As shown, incoming RF signals are received via an antenna 510 and provided to an RF front end circuit 515. Front end circuit 515 may perform various RF processing, including amplification in a low noise amplifier, filtering, downconversion and digitization. In turn, resulting digital information, e.g., at baseband, is provided to a baseband processor 520.
Baseband processor 520 may perform various preparatory processing before providing resulting SME's to a SWAM demodulator 530. As shown, demodulator 530 is configured to perform a reverse set of operations as performed in a modulator on the transmit side. The selection of operations and configuration of demodulator 530 may be performed under control of a controller 550, which may determine an appropriate configuration of demodulator 530 to perform demodulation based at least in part on the information parameters, including Range, MCS information and Group information, as discussed above. Depending on implementation, this information may be received as one or more of network information, feedback information, and/or measurement information. Based on at least some of the information parameters, controller 550 may determine an appropriate reverse Recipe of operations to perform on incoming SME block to enable recovery of a given symbol as well as to determine whether shaping operations including edge processing, window and filtering or so forth is to occur. Controller 550 may be a programmable hardware circuit that is configured to execute instructions such as may be stored in one or more non-transitory storage media.
As shown, SME blocks are provided to an optional edge processor 532 to perform any edge processing, windowing or filtering as indicated under control of controller 550. Thereafter the SME blocks are provided to a Recipe operator 534, which may perform a reverse Recipe of operations on the SME block, as instructed by controller 550 to recover original SSs. Thereafter, via combiner 535 the separate SSs are aggregated into a single denoised SS.
In turn, this SS is provided to a FFT engine 540 for transformation to the frequency domain. Understand while shown in
The frequency domain SS (in the form of a plurality of subcarriers) is provided to a correlator 542 for execution of a correlation as a dot product operation between an incoming SS and the conjugate of selected SS 554 (transformed into the frequency domain via FFT engine 555). Note that controller 550 may select selected SS 554 from a pool of SS 552 based on the information parameters (e.g., Range, Group).
Still with reference to
In turn, this symbol is provided to a symbol-to-bit decoder 560 to output information bits that are then provided to an optional descrambler 570. Descrambler 570 may descramble the information bits to result in a bitstream that can be provided for further processing within baseband processor 520 or another location within a receiver. Understand while shown at this high level of
Referring now to
Still with reference to
Next at block 650 the determined reverse Recipe of operations may be performed on modulated symbols of the SME. An aggregation process then is performed at block 660 to aggregate the processed modulated symbols to result in a single SS. Note that the processed modulated symbols may present different levels of reliability (or equivalent SINR), especially if interleaving options are activated (sub-blocks spread in time and/or frequency); then, advanced receivers could detect/measure this to weight the processed modulated symbols before aggregating them to a single SS.
Thereafter at block 670 a correlation may be performed between the single SS and a selected SS. In this way, soft estimations for the symbol are obtained, also termed “symbol-likelihood-values,” as in some cases a soft decision analysis may be more optimal to provide more/accurate soft information to a symbol-to-bit decoder (such as decoder 560 of
Then it may be determined at diamond 670 whether more symbols are present. If so, control passes back to block 630, discussed above. After multiple symbols have been obtained, control passes to block 680 where these soft estimations may be processed to obtain a set of information bits. In this process, given the spreading of information over multiple SMEs, several soft decisions from several symbols may be processed to properly deinterleave (if needed) and decode (in an FEC decoder) them to obtain the final information bits. Thus in the embodiment of
In one embodiment, the correlator and analyzer processing may be done in time from the received and expected sequences. One or several correlations or part of correlations (per block), with one or several circular shift and filtering of one of the sequences (received or expected, one of them being generally conjugated before the operation), provide enough information to the analyzer to perform the symbol detection or the delivery of (generally 2M) symbol-likelihood-values for each received SME. These symbol-likelihood-values can be seen as real values measuring the probability that the received symbol is equivalent to each of the 2M potential symbols; the higher is the better. The soft-decoding (reverse FEC) can advantageously benefit from this information to provide the most probable bitstream.
In another embodiment, the correlator processing may be done in frequency from the received and expected sequences. One or several correlations, dot product or part of it, with one or several circular shift and/or filtering of one of the sequences (received or expected, Fourier-transformed, one of them being generally conjugated before the operation), provide enough information to the analyzer to perform the symbol detection or the delivery of (generally 2M) symbol-likelihood-values for each received SME.
In another embodiment, the correlator processing may be done in frequency and in time from the received and expected sequences. One or several correlations, dot product or part of it, with one or several circular shift and/or filtering of one of the sequences (received or expected, Fourier-transformed, one of them being generally conjugated before the operation), can be done to estimate the most probable modulation shifts. Then, another processing, in frequency or in time, from the same received and expected sequences with additional processing (circular shifts, filtering), can refine the estimation of correlation probabilities around the expected shifts (up to 2M potential symbols) to provide enough information to the analyzer to perform the symbol detection or the delivery of (generally 2M) symbol-likelihood-values for each received SME. This extra processing can also be used to track the TX-to-RX drift, i.e., synchronization derives, to correct the receiver algorithms and/or potentially adjust its clocks.
In one embodiment, as a simple and robust solution, the analyzer can search for most probable shifts or symbol-likelihood-values using the absolute values of the FFT of correlation(s) (without considering the phase). This is a robust solution to provide values without needing accurate synchronization nor tracking.
In another embodiment, as an algorithm improvement, the analyzer can consider the complex-valued FFT/correlation(s), and potentially previous phase estimation, to decrease its equivalent noise in correlation (projection on a given phase of the noise can divide by a factor 2 the equivalent noise after correlation).
In another embodiment, as another algorithm improvement, the analyzer can consider several times the previous processing, like a Rake-receiver, after the estimation of several relevant paths (CIR) in a channel with strong echoes. Such a multipath environment detected in previous synchronization steps or from the tracking process, can trigger this extra processing to still improve performance in these most difficult channels.
Note that in case of linear-sweep SS, optimizations for efficient demodulation can be done replacing intensive Rake processing by simpler filtering, as the CIR appears in correlations steps; allowing several designs of low-power low-cost receivers with good performance.
To improve performance while keeping a rather low processing complexity, the Maximum-Ratio-Combining principle can then be applied on main channel paths (CIR), but also on the recombination/aggregation of subblocks of SMEs especially if interleaving principles are considered in the modulation. Over the whole duration and bandwidth of the SME, if SME is split in P subblocks in time (see interleaving of
In one embodiment, to further improve performance (including increase coverage while preserving the wireless resource limiting retransmissions, also allowing measurements among the gateways to select the best one(s) for each user), in a network where several Gateways receive the transmitted packet in uplink transmissions, the symbol-likelihood-values for each received SME of a given packet can be generated at each Gateway. These values are then sent to a centralized server that will optimally gather and decode this soft-information to provide the final bitstream.
Embodiments may be incorporated into many different types of wireless systems. Referring now to
In the embodiment shown, integrated circuit 700 includes a memory system 710 which in an embodiment may include a non-volatile memory such as a flash memory and volatile storage, such as RAM. In an embodiment, this non-volatile memory may be implemented as a non-transitory storage medium that can store instructions and data. Such non-volatile memory may store instructions, including instructions for transmitting and receiving packets having SWAM modulation characteristics as described herein.
Memory system 710 couples via a bus 750 to a digital core 720, which may include one or more cores and/or microcontrollers that act as a main processing unit of the integrated circuit. In turn, digital core 720 may couple to clock generators 730 which may provide one or more phase locked loops or other clock generator circuitry to generate various clocks for use by circuitry of the IC.
As further illustrated, IC 700 further includes power circuitry 740, which may include one or more voltage regulators. Additional circuitry may optionally be present depending on particular implementation to provide various functionality and interaction with external devices. Such circuitry may include interface circuitry 760 which may provide interface with various off-chip devices, sensor circuitry 770 which may include various on-chip sensors including digital and analog sensors to sense desired signals, such as for a metering application or so forth.
In addition as shown in
Note that ICs such as described herein may be implemented in a variety of different devices such as an IoT device. This IoT device may be a smart utility meter for use in a smart utility network, e.g., a network of stars or gateways in which long-range communication is according to an IEEE 802.15.4 specification or other such wireless protocol.
Referring now to
As an example, RANGE0 and GROUP0 parameters could be fixed between end node device and gateway (with a default highest range mode and group 0 at initialization/reset, and default values known after the device is registered and associated with this network). RANGEO could have 4 sequence sizes for instance. GROUPO could use the sequence in one direction or in the other (i.e., apply a conjugate operator or not); band selection also extends these groups. Gateways listen to all RANGE0 and GROUP0 modes, enabling quasi-orthogonal simultaneous packets.
An end node device (e.g., one of IoT devices 810) generally listens only for its selected RANGE0 and GROUPO0DL mode in its assigned band, at given instants, to minimize complexity and power usage. Then, RANGE1 and GROUP1 parameters may also be defined, but be more variable and defined at each transmitted packet to scale and adapt. RANGE1 could be defined from a preamble and/or header to help in further scaling the packet in range, i.e., length/rate (to refine range and MCS). GROUP1 could extend the distinction between networks/services/IDs/groups to help in further scaling in density, e.g., limiting the interference between different packets (of different networks), and stopping receiver processing as soon as possible if the packet need not be decoded in a given receiver.
In various embodiments, different SSs and Recipes can be used not only for each frame (depending on frame and group) but also for each SME index (generally looping over a given sub-list of SSs and Recipes as presented above). The main goal of such variations is to mitigate the interferences between similar SWAM packets. Instead without an embodiment where there are no variations (typically the case in conventional standards and long-range solutions), several packets sent at the same time would interfere and impair receipt, such that a global system could quickly go down with an escalation of retransmissions.
Embodiments thus provide flexible modulation/demodulation that is scalable (and customized to cope with given applications without compromising simple and efficient reception, keeping the expected properties), providing sufficient degrees of freedom and algorithms to exploit these signal properties.
With the aforementioned design choices, as throughput mainly depends on Range and MCS parameters (sequence length, SS to SME repetitions, number of shift states per SS, FEC coding scheme) and PIM capabilities depend on other parameters (number of operation combinations per SME over the SS to SME repetitions, SS list size, potential extra SME basis rotation), the system is rather flexible and scalable to adapt to many long-range applications. Trade-offs also are possible to increase either user throughput or robustness with many simultaneous users.
In a long-range network, packet transmission may occur with a random access, i.e., unslotted/unmanaged, such that any device can send non-synchronized packets, e.g., in an uplink direction. Embodiments enable scalability in packet generation to be efficient on a wide range of scenarios.
In some embodiments, packet structure may be similar to a IEEE 802.15.4g Wireless Smart Utility Network (WiSUN)-orthogonal frequency division multiplexing (OFDM)-based packet, with variations to ensure scalability.
In an embodiment, the packet is formed with: a preamble having a variable length that is used for detection, where the length scales with a generally repeating known sequence; next further training fields allow finer synchronization (in frequency and time) and also blindly deliver information to detect a range mode to scale the packet accordingly; thereafter scalable header fields may be split into multiple parts to optimize packet duration and start mitigating packet interference as soon as possible, with several Groups so that several concurrent packets do not interfere too much (also enabling to stop demodulation as soon as possible, if decoding fails or user group is different); and finally a data portion is provided that also may have a variable length.
This progressive scalability in rate and in groups to reduce interference enables efficient reception of the data portion of the packet for a targeted application.
Each frame part is defined from 3 parameters {Range R, MCS M and Group G}. The progressive refinement of these parameters over a packet is a key to efficiently scale and preserve robustness. Each packet portion (e.g., PHR1, PHR2, DATA) is formed of a plurality of SME symbols, where each SME is constructed from an SS sequence and Recipes of operations.
Consider an example of a high-level choice to customize settings based on network requirements and environment, allowing to define for each packet these 3 parameters of (Range R, MCS M and Group G). These 3 parameters enable customization of packets and to adapt SWAM modulation with a given list of SSs (all SSs may be configured to have the same length for a given Range), given lists of Recipes (customized from MCS and Group), and with given potential variations of the Recipes over the frame to further improve packet-interference mitigation.
Just to provide an example of packet structure and duration for the main Ranges RGS, Table 5 below shows example customizations. As shown, in a 500 kHz Bandwidth configuration, shortest range signals could be based on SUN-OFDM option2, and longest-range signals target the worst-case with a sensitivity down to −130 dBm and large channel echos. This long-range case corresponds to lowest data rates of 400 bits per second; while highest rates can reach SUN-OFDM data rates with tens of kilobits per second.
This flexibility to ensure a scalable system, customized to the network needs, to balance the efficiency in terms of Range, throughput, number of simultaneous users per gateway, and so forth. Robustness is enabled using SWAM modulation and configurable packets in accordance with an embodiment.
Referring now to
While described as “transmitter” and “receiver,” understand that the operations performed on this transmit side can be performed by gateways, access points or so forth, as well as stations. In any case, understand that both transmitters and receivers may be part of devices that have transceiver functionality and thus, both can transmit and receive information.
As illustrated, method 900 begins by identifying a range for a device pair, namely a pair of devices in communication, including a transmitter and a corresponding receiver (block 910). Depending upon implementation, this range may be identified based on previous link parameters. Still further, the range may be adapted if necessary, e.g., over multiple packets. Thereafter at block 920, a preamble is generated for the packet. This preamble may have a length based at least in part on the identified range.
Still referring to
Thereafter at block 940, the data portion of the packet may be generated according to various information present in one or more of the first, second and third fields. Next at block 950, the packet is transmitted (to the receiver(s)) with all of the preamble, header and data portions.
Still with reference to
Otherwise, when it is determined to update parameters, control passes to block 980 where such one or more parameters may be updated based on the exchanges between devices. Such exchanges may include measurement information, feedback information, communication of packets indicating modification to the parameters or so forth. Understand while shown at this high level in the embodiment of
Referring now to
As illustrated, method 1000 begins by receiving a preamble of the packet in the receiver (block 1010). Next at block 1020, correlations may be performed to detect the preamble. In an embodiment, time shifting correlations may be performed on the preamble, which as discussed above includes repetitions of a sequence, enabling simple but efficient packet detection. A similar processing in the frequency domain (using FFT) is also possible.
Next at block 1030, the detection may be validated based on the end of preamble. In an embodiment, this detection process may include multiple correlations with varying preamble signals (including signal inversions, conjugation, etc.) or other known operations on the sequence repetitions to validate the signal in a noisy environment. Thereafter at block 1040, the receiver may be synchronized and a range mode identified, e.g., based on a first field of the header. Next, the range mode may be refined (optionally) and a group identified based on a second field of the header (block 1050). Still further, at block 1050, an MCS definition may optionally be refined for the following fields.
Next at block 1060, various parameters for a data portion of the packet may be determined based on first, second and third fields of the header. For example, this information may identify a data length and an optional refinement of MCS definition for the data portion. Finally at block 1070, data of the data portion may be demodulated based at least in part on these parameters.
Referring now to
As illustrated, packet 1100 includes a preamble 1110 that is formed of two fields, a first variable-length training field 1112 (STF1) that includes a repetitive sequence to enable a low power packet detection and a second (potentially) variable-length training field 1114 (STF2) that provides a signature for detection validation (and potential filtering to stop packet processing if the signature does not match the receiver expectations).
Packet 1100 further includes a header 1120 that can include a first variable length field 1122 that is configured as a long training field used for synchronization refinement and to potentially provide range information refinement, a second variable length field 1124 that is configured as a first packet header field (PHR1) to provide top mode signaling including group information, potentially additional range information or refinements, and a second variable length field 1126 that is configured as a second packet header field (PHR2) to provide additional signaling information, including MCS and/or length information for a data field 1130, which includes variable length data. Note that data field 1130 may include a given payload. Understand while shown with a particular implementation in
For example, a preamble could be used not only for packet detection, validation and synchronization, but also to filter/identify a given network/service/ID. In this way, a preamble may directly provide a way for an early break (to continue receiver processing only if the packet is intended for a valid network/service). The preamble may also provide rough initial range (and maybe group/ID) information too, before the header.
Then, there can be two options to provide main Range bits, but not exclusive (progressive refinement). For instance, a part of the scaling/range, high-level, could be fixed/pre-negotiated (e.g., at device association on the network), defining the sequence length and RGS info as a quasi-fixed definition from a beginning of a packet, (e.g., directly from the beginning of a preamble); and/or another part (finer granularity of range) could be defined in the preamble as a pool of signatures (e.g., double the number of ranges, adding 2 signatures per range that the receiver will be able to distinguish) and/or in header fields (fields to still refine granularity of range). The same split of information for IDs (or groups of other information to identify/split among networks or users). For example, a network/service ID at a preamble validation process could avoid further decoding of packets that are not for a given network ID that receivers expect given signatures among a pool. Such information could be provided at an end of a preamble and/or a beginning of one or more header fields.
Of course in other implementations, there can be different or additional fields within a packet and other scalability features are possible.
In embodiments there may be additional system choices for the various packet fields. In some cases, the STF1 and STF2 fields may be fixed for a selected network configuration. However, as STF1 is a repetition of recognizable sequences designed to target a given sensitivity, its length can be updated to match a range/sensitivity dimensioning of a current packet (e.g., adding some repetitions to target a lower sensitivity, to allow a receiver to detect it in a more difficult and noisy channel at the expense of a longer packet) without altering the algorithms. In other cases, STF1 and STF2 fields could have a small power boost factor (and/or rescaling factor by reducing the sampling rate, that would simultaneously reduce the bandwidth and increase the duration), to improve performance while still being within regulatory constraints, e.g., if these parts have a lower PAPR under a maximum-power constraint. As the LTF/PHR1/PHR2/DATA fields can be based on the same SS sequences discussed above, any or all can be used at a receiver to improve synchronization.
One example of dimensioning technique is now described. First, a LTF can be implemented as a non-modulated PHR1, plus a simple specific coding based on simple inversion of real and/or imaginary parts of the SS (=conjugate=inverse-sweep-orthogonal sequence). These coding options allow delivering 2 bits of information for a main Range, to scale the packet structure (and potentially use this first level information as a top Group-ID1 for PHR1 and subsequent packet fields). This assumes a blind detection at receiver of these 22=4 potential sequences to select these 2 main bits of RanGe Scheme mode (RGS). These 4 modes also correspond to (up to) 4 LTF lengths to optimize synchronization vs. range. This arrangement also allows early stop (back to sleep) of decoding if the packet is not for the user (if we consider 4 modes, there is a 75% drop probability), for power savings. It assigns one of the 4 MCSp1 (as shown in Table 5, where MCSp defines MCS for PHR1) for the following PHR1.
Lower Range modes assume short SS sequences, while longer Range modes use a standard SS (e.g., ‘twice-length’). Once this SS length is defined, the same SS length may be assumed over the packet, and subsequent signaling will play on other parameters and MCS to balance range/performance.
PHR1 contains information to refine the Range mode/MCS for the following parts and the Group information. In an embodiment, PHR1 contains 5 bits: 1 bit used to refine the Range/MCS list for PHR2 and DATA, and 4 bits to refine the Group-ID2. In other words, the robust PHR1 contains top-level budget public information to define 8 MCSq (where MCSq defines MCS for PHR2) (3bits: 2 from LTF and 1 from PHR1) for PHR2 to further scale and optimize length vs robustness; and a GroupID information from this signaling is used to select the Recipe to construct SMEs (and improve Packet-Interference-Mitigation), and also potentially update the scrambling of binary content.
PHR2 can potentially be modulated at higher throughput (among the 8 MCSq), and contains remaining signaling information. For example, it can contain MCS data details for DATA, Data FrameLength, options for framing, validation, addressing or other packet control.
In an embodiment, PHR1 and PHR2 are block-encoded. Such forward error correction (FEC) allows efficient adaptation of the number of input information-bits to the number of coded-bits to fit the number of dedicated SMEs (in particular to be a multiple of the number of bits per SME). Moreover, such coding schemes on small blocks can efficiently consider trade-offs with embedded detection of errors and correction, potentially removing the need of costly (long) CRC for validation, and allowing a process stop if the packet is not relevant (reliable/decodable) for the user.
Efficient FEC and bit-interleaving/mapping, optimized for each part (PHR1/PHR2), can be considered, with good capacity of error detection and correction (e.g., ensure an error misdetection probability <10−5) while allowing one or more bits to be corrected for each PHR part, with adapted brute-decoding for small PHR1 and optimized-iterative decoding for PHR2.
A data portion may be modulated as defined in the final MCS combination from previous signaling. In one embodiment, 8 MCS modes per Range mode may be used.
Note that additional improvements are possible (e.g., merging LTF with either end of preamble or PHR1, reducing LTF length performing sync-timing on PHR1 content with potential iterative demodulation, header compression, etc.).
The following details are provided for the STF part (also referred to as PDF for “Packet-Detection-Field”). Depending on implementation, several solutions are possible. To integrate existing standards as an add-on, the first detection part could be similar to the considered standard (e.g., IEEE 802.15.4g WiSun STF). To improve detection and synchronization for Long-Range systems, the sequences can be improved, e.g., considering non-equally spread tones in frequency (and optionally evolving content over STF1 and STF2, especially to allow variable-length STF1 while improving synchronization capabilities in the STF2 part), with well-chosen complex phase values to reduce PAPR.
To further capitalize on frequency-sweep SWAM-like signal shapes, the preamble could be based on such frequency sweeps (SS), or combinations of frequency sweeps, to allow easy/efficient packet detection while keeping PAPR very-low (additional power compression and power boost is also possible). A given signature can then validate the packet start, and could also select sub-groups to minimize unneeded demodulation. Moreover, this could allow use of the same algorithm on a larger part of the packet, to “merge” packet detection or synchronization parts with signaling parts (e.g., replaying the recorded signal at receiver if needed).
Then, it can be merged at least with LTF. Synchronization on such sweeping sequences (sweep at sample-level and repetition at block level) allows rather simple (first combination/sum to reduce noise, sweep tricks for combined-sync time/frequency) but efficient and scalable fine time and frequency synchronization.
In one example, a preamble can have irregular active-subcarrier spacing and frequency-shift at a block-level. One example preamble could be to use a ‘zigzag’ STF1 (using a given default preamble (SSp) with low PAPR, with a STF1 formed of repetitions of this SSp and time-reversed or conjugate of SSp), with sufficient repetitions for the targeted range; and a ‘signature’ STF2, which validates that a packet from this standard is detected, and improves synchronization in time and frequency (removing potential ambiguities).
Note that the initial part (STF1 repeated pattern-signals) can be scalable in length (e.g., number of repetitions) to respect the advised principle of homogeneous scalability over the packet for a given targeted range. Increasing the STF1 length allows receivers to hang more long-range packets (increase detection probability, targeting lower SNR/sensitivity), for worst-case ranges. Examples of global STF duration for each Range is provided in Table 5 above (8 ranges for 8 lengths).
The following details are provided for the LTF part (also referred to as VTF for “Variable-Training-Field”). In an embodiment, the LTF can use the same SS as PHR1. As an example, to provide the 2 bits of RanGe Scheme mode (RGS), 4 types of LTF are defined, with potential relations between them to easily and blindly detect these modes at receivers. Typically, 2 SS and/or simple operations of inversion (conjugation) on real and imaginary parts of portions of the signal (like is done with SWAM recipes on SS) allow simple and efficient (including low-power) detection of these SS-based signals with good properties, and efficient synchronization at receivers. As the 4 modes correspond to 4 targeted ranges (and related SNR levels), longer-range modes will have longer LTF, the LTF SS repetitions/recipes lengths can be customized to this range mode to enhance efficiency with well-calibrated packet parameters and lengths (homogeneous scalability is key; 4 RGS in LTF=4 lengths). Exit conditions can be customized also.
Note that at a receiver, a channel profile (CIR) could also be deduced from the preamble+LTF processing. This could help algorithm customization in the receiver, and could also be potentially reported for subsequent MCS choices in future packets (enabling high-level control, e.g., by adjustment of the MinShift parameter to set the number of bits per SME symbol). Also note that at the receiver, for synchronization refinement, algorithms could consider STF/LTF, and the following parts (like PHR1) as they are based on the same kind of sequences (e.g., once demodulated PHR1, adjust sync, to increase robustness for PHR2 and subsequent signal).
The following details are provided for the PHR1 part (“Packet-HeadeR 1”). The PHR1 contains main header information, to further scale the subsequent packet signal (PHR2). Consequently, only a few bits are provided in this robust part to be efficient. A main refinement of Range/MCS and Group can be given here. In an embodiment, the robust PHR1 containing top-level public information may be provided using 5 bits: 1 more bit to be used to refine the Range/MCS list (for subsequent PHR2 and DATA, to enhance the granularity of rate vs error-probability); then it will allow selection from among 8 defined MCSq (3bits: 2 from LTF and 1 from PHR1) for PHR2 to further scale and optimize length vs robustness; and 4 bits to define the Group-ID2 (for subsequent PHR2 and DATA, to enhance the robustness vs interferers and similar packets). This part also allows selection of the Recipes to construct SMEs (and improve Packet-Interference-Mitigation) for next parts, and also potentially update the scrambling of binary content.
PHR1 follows the principle of scalability, like LTF, with an MCSp (and then duration) adapted to the range of the packet (4 RGS from LTF=4 lengths). Then, in an example, up to length (RGS)=4 MCSp can be defined and used for PHR1. Examples of proposed MCSp {0,1,2,3} are shown in Table 5 above. Good block-encoding like BCH (15+X,5) may be used, where X={0,1} for instance, depending on MCS to match symbol constraint (to be multiple of 2 or 3 bits).
The following details are provided for the PHR2 part (“Packet-HeadeR 2”). The PHR2 contains additional header information, to further scale the subsequent packet signal (DATA). Consequently, more bits are provided to refine MCS and Group for DATA part, and also to provide useful information like DATA length (e.g., number of bits or bytes or non-uniform length mapping to save bits in this header while covering a fairly-wide range of packet sizes), and other packet and system information (e.g., one or more of TX-power, Version, TRX mode, Retry mode, Security options, CRC mode, ACK mode, or so forth). In one embodiment, the PHR2 may contain approximately 16 bits, as follows: 3 more bits to refine the MCSdata to define 8 potential MCS per Range, for DATA transmission to further scale and optimize length vs robustness; 7 to 8 bits to set Frame-Length (could depend on the MCSq mode); 2 to 3 bits to set TX-power; and 3 or more bits for Packet and System information (e.g., modes for
Version/TRXmode/Security/CRCmode/Retry, etc.). PHR2 follows the principle of scalability, like LTF/PHR1, with an MCSq (and then duration) adapted to the range of the packet (8 Ranges from LTF+PHR1: 8 lengths). Then, in an example, up to length (Range)=8 MCSq can be defined and used for PHR2. This field may use good block-encoding like BCH (30+Y,16), where Y={0,1,2,3 . . . } depending on MCS to match symbol constraint to be multiple of 2 or 3 bits, and also customize robustness.
The following details are provided on the DATA part (also called “PSDU” content). The Data part is scalable in number of bits contained (defined from previous field FrameLength in PHR2) and in MCS (from the 8 MCSdata defined for each of the 8 Ranges). The PSDU payload includes a MAC header (variable with more fields), Data content (and potential tail/padding) and CRC (options). Note a GroupID and Security mode may also impact the transmitted signal. Then, a data portion from approximately 1 ms to more than 1 second could be defined depending on the parameters.
In various embodiments, system-wide high-level smart management of packet parameters may occur to optimize communications with one or more devices considering all of range, users in the environment and feedback information regarding prior communications.
In one or more embodiments, the Range parameter can be defined from fixed/expected channel characteristics between transmitter and receiver (including, for example, distance/echos/SNR) and/or refined from previous information/exchanges (such as RSSI of previous packets, rough estimation of SNR/RSSI at commissioning, or so forth). Even for a given parameter (e.g., Range), a part can be implicit (or fixed by the network or at network registering/association or any previous exchange) and another part explicitly coded in the packet. Note that error rates and retransmissions could also be used to adjust following packets' range level.
In one or more embodiments, the Group parameter can be selected randomly, and/or from UserID/address information, and/or from a computation or negotiation with the network for a given user or transmission. In selecting the GroupID, goals of maximizing global capacity and minimizing probability of interfering packets may be considered.
System-level management can improve selection/adaptation for each user. Moreover, in a star system with Gateways (GWs) that can be more complex with power and space adaptation per user packet, beams can be adapted to reduce interferences (path angles, powers, MCS, sequences and orthogonality management for packets in closer areas) on a given band. Specific subgroups (of sequences) quasi-orthogonal to other can be defined to ensure better behavior (and orthogonality) for specific packets (first or specific access, control, quality of service (QoS) or so forth). As an example, embodiments may define certain modes/groups with better characteristics (e.g., best SS and Recipe combination) for highest priority packets.
Embodiments may also leverage flexible packet definitions as described herein to perform link adaptation, providing coherent and flexibility using the modulation described herein (where each device packet can have different targets, budget/sensitivity/range, rate, capabilities and so forth). As discussed above, various parameters can provide degrees of freedom to optimize wireless links. Such degrees of freedom may include: SWAM modulation based on control of one or more of Range, MCS and Group information; header/control updates based on link information including power, data lengths, channel band-like information, addressing; security, application and MAC-level information.
Depending on use case, there may be several levels of parameter adaptation (e.g., fixed, slow, quick). There can be different ways to exchange/adapt these parameters. In some cases, parameters can be controlled by: fixed pre-negotiation or defined at association; measures including channel quality (e.g., LQI/RSSI/SNR/CIR); exchanges with information in each packet/header or dedicated control); and/or receiver a priori knowledge vs blind detection.
With sub-group assignments, there can be dedicated groups for specific packets. For example, a special configuration may be provided for robust/specific packets. Or there can be a smart allocation of channel bands and groups in each band for different associated devices (packet interference mitigation optimization for the network system), enabled using flexible SWAM modulation in accordance with an embodiment.
Understand Embodiments may also enable high-level error mitigation. For example, packets may be retried with different group information. Such variations may mitigate signal issues in the given channel. In other cases, retried packets may be sent using constant group information, which may allow reliability aggregation of received packets to recombine soft information at an application server to properly decode the data.
