Patent Application
20090253376
The teachings herein relate generally to wireless networks and devices operating in such networks, and are particularly related to sensing spectrum used in the network such as to avoid interference or otherwise manage spectrum usage.
Spectrum sensing is needed in cognitive radios to find empty slots in the radio spectrum which can subsequently be used in an opportunistic manner. Traditionally radio spectrum is divided between different radio systems in a manner that strictly allocates a specific band to a specific system. This strict allocation will be changing to a more flexible spectrum utilization at least in some frequency bands in the future. Primary users are those operating within the more formal networks such as hierarchical networks (e.g., WLAN or cellular such as GSM, GERAN, UTRAN and E-UTRAN) and ad hoc networks (e.g., WiFi). Secondary users are those operating outside the structure of the formal networks. Since essentially all spectrum in crowded areas that is useable by mobile terminals is allocated to some formal network or another, the secondary users find and utilize portions of the existing networks' spectrum in an opportunistic manner. Consequently, two related obstacles face the secondary user: it must not interfere with the primary users, and it must somehow find those portions of the spectrum not currently in use by any of the formal networks. For this latter reason the secondary users are generally referred to as cognitive users; they must be spectrum-aware rather than simply using the radio resources allocated by some access node controlling a cell of users.
The secondary user/cognitive radio therefore utilizes or exploits a free region of spectrum for its own transmissions, outside control of the formal networks. By “free” is meant that the primary users/formal networks are not using the spectrum region in question when considering time, frequency and space. Alternatively there could be a band that is dedicated to several radio systems operating under a certain set of rules or policies. The common factor in any case is that the radio spectrum will have to be sensed somehow in order for the cognitive radio/secondary user to locate the free spectral band. This sensing has to at least take into account time, frequency and space.
The cognitive radio must be able to detect a primary user and avoid causing interference to it. Since cognitive radio is a wireless application, the cognitive radios have a limited power supply and so an important consideration in spectrum sensing is to minimize power usage. It is not inconsistent that the actual mobile terminals operating within the formal network(s) may act as the cognitive radios that also sense the spectrum quality, but the spectrum sensing task may also be undertaken by the cognitive radios operating wholly outside those networks' formal structure but utilizing the free radio resources opportunistically.
In theory the spectrum sensing task can be shared among various cognitive users in order to obtain savings in power consumption at the individual mobile users, to reduce the amount of data to be transmitted by those individual users, and consequently to facilitate the use of simpler sensing techniques in each individual user as compared to having individual terminals measure the entire spectrum in use across the network. One practical problem with collaboration is that the selection, control, and resultant handling of the collaborative nodes represent a control signaling overhead that occupies additional bandwidth because the collaborative nodes need to be controlled independently. Thus the potential spectrum savings by allocating bandwidth more efficiently in view of the spectrum sensing by the mobile terminals is offset, often more than offset, by the coordination required to get those mobile terminals to sense portions of the spectrum. This is particularly true when one tries to incorporate diversity into spectrum sensing so as to avoid the hidden node problem. Teachings relevant to collaborative spectrum sensing in a cognitive network may be seen more particularly at co-owned U.S. patent application Ser. No. 12/001,623, filed on Dec. 11, 2007 and entitled “Method and Apparatus to Select Collaborating Users in Spectrum Sensing”.
Being a very forward-looking technology at this stage of development, there is not a great volume of prior art in the spectrum sensing field. Three are detailed here. In a first proposal set forth by L. Luu and B. Daneshrad in a paper entitled An Adaptive Weaver Architecture Radio With Spectrum Sensing Capabilities to Relax RF Component Requirements [IEEE J
A second paper by J.Laskar, et al, entitled Reconfigurable RFICs and Modules for Cognitive Radio [SIRF, 2006] presents a sensing method which is based on a coarse detection with wavelet transformation and then on a fine detection that uses an analog auto-correlation. It may in principle be suitable for the spectrum sensing task for cognitive radio applications at least with some properties. Few details are presented though and so it is difficult to find a workable solution for the specific cognitive radio problem. For example, this paper does not address the time-division of signals under detection in any way.
A third approach was presented by J. Yang, entitled Addressing the dynamic range problem in cognitive radios [Berkeley Wireless Research Center (BWRC) summer retreat, May 31, 2007, published in ICC conference, July 2007] proposed a search for the highest amplitude peak over the frequency from the wideband spectrum with a low resolution, high-speed analog to digital converter ADC and correlator. This proposal then cancels the result from the original signal before a second digital conversion that uses also the high-speed, low resolution ADC. It appears to the inventors that the dynamic range for the second ADC is significantly reduced in this proposal, though the wide-band signal may still be converted completely.
What is needed in the art is a way to find those free areas that may be located anywhere among the wideband spectrum at various times with low power requirements and high confidence level.
In accordance with one embodiment of the invention there is a method that includes searching a wideband spectrum with a first (lower) resolution to identify bandwidth slices having low or no signal levels and searching the identified bandwidth slices with a second (higher) resolution and selecting as candidate frequency slices those bandwidth slices having the least signal levels after the second resolution searching. The method further includes executing a spectrum detection algorithm on at least one of the selected candidate frequency slices and deciding that it has sufficiently free spectrum, and then transmitting in the decided candidate frequency slice.
In accordance with another embodiment of the invention there is an apparatus that includes a coarse power detector, a fine power detector, a spectrum detector, a processor and a transmitter. The coarse power detector is configured to search a wideband spectrum with a first (lower) resolution to identify bandwidth slices having low or no signal levels. The fine power detector is configured to search the identified bandwidth slices with a second (higher) resolution and to select as candidate frequency slices those bandwidth slices having the least signal levels after the second resolution searching. The spectrum detector is configured to execute a spectrum detection algorithm on at least one of the selected candidate frequency slices. The processor is configured to decide that at least one of the selected candidate frequency slices has sufficiently free spectrum. And the transmitter is configured to transmit in the decided candidate frequency slice.
In accordance with another embodiment of the invention there is a computer readable memory embodying a program of machine-readable instructions executable by a digital data processor to perform actions directed toward searching spectrum for a frequency band available for transmission. In this embodiment the actions include searching a wideband spectrum with a first (lower) resolution to identify bandwidth slices having low or no signal levels and searching the identified bandwidth slices with a second (higher) resolution and selecting as candidate frequency slices those bandwidth slices having the least signal levels after the second resolution searching, and thereafter executing a spectrum detection algorithm on at least one of the selected candidate frequency slices and deciding that it has sufficiently free spectrum.
In accordance with another embodiment of the invention there is an apparatus that includes first searching means for searching a wideband spectrum with a first (lower) resolution to identify bandwidth slices having low or no signal levels, second searching means for searching the identified bandwidth slices with a second (higher) resolution and selecting as candidate frequency slices those bandwidth slices having the least signal levels after the second resolution searching, detect means for detecting spectrum of at least one of the selected candidate frequency slices, processing means for deciding that at least one of the selected candidate frequency slices has sufficiently free spectrum, and transmit means for transmitting in the decided candidate frequency slice. In a particular embodiment, the first searching means is coarse detection circuitry 304, the second searching means is fine detection circuitry 306, the detect means is an embodied spectrum sensing block 308 that runs a spectrum detection algorithm, and the transmit means is a transmitter 10D.
These and other aspects of the invention are detailed more particularly below.
The foregoing and other aspects of these teachings are made more evident in the following Detailed Description, when read in conjunction with the attached Drawing Figures.
The inventors' solution to the problem formulated above, in an embodiment, breaks the spectrum sensing operation into three main parts: coarse wideband spectrum sensing; narrowing the band and doing a power analysis of that narrower band to find a candidate band; and finally analyzing the selected candidate band for free spectrum. In this embodiment there is also a low power envelope detector constantly operating in the background to detect signals which emerge at long intervals. This envelope detector then gives information that triggers spectrum sensing when necessary to detect those infrequently appearing signals.
The teachings herein detail sensing the surrounding radio spectrum and how to find a free slot from there. As noted above, an important requirement in mobile applications (in addition to the basic functionality) is the power consumption. These teachings are in the context of spectrum sensing and finding a free slot in a power-efficient manner, and also in an approach that is suitable for mobile terminals. The inventors have determined that a significant bottleneck in the mobile terminals lies in the analog to digital A/D conversion. It is at this juncture where the engineering tradeoffs among speed, resolution and power will not allow a purely digital solution with the current or foreseeable technologies that are suitable for small (handheld) mobile devices. This is not to say that these teachings are limited only to handheld mobile devices, but only that they are particularly advantageous for them.
Prior to detailing the various embodiments and aspects of the invention further, some general information as to the cognitive radio and its environment are presented at
The terms “connected,” “coupled,” or any variant thereof, mean any connection or coupling, either direct or indirect, between two or more elements, and may encompass the presence of one or more intermediate elements between two elements that are “connected” or “coupled” together. The coupling or connection between the elements can be physical, logical, or a combination thereof. As employed herein two elements may be considered to be “connected” or “coupled” together by the use of one or more wires, cables and printed electrical connections, as well as by the use of electromagnetic energy, such as electromagnetic energy having wavelengths in the radio frequency region, the microwave region and the optical (both visible and invisible) region, as non-limiting examples.
The PROG 10C is assumed to include program instructions that, when executed by the DP 10A, enable the cognitive radio terminal 10 to operate in accordance with the exemplary embodiments of this invention, as detailed above. Inherent in the DP 10A is a local clock to enable synchronism among the various terminals, which is important in some cognitive radio architectures. The PROG 10C may be embodied in software, firmware and/or hardware, as is appropriate. In general, the exemplary embodiments of this invention may be implemented by computer software stored in the MEM 10B and executable by the DP 10A of the terminal 10, or by hardware (e.g., ASIC10E or other firmware circuitry), or by a combination (e.g., FPGA 10E) of software and/or firmware and hardware in the terminal 10.
In general, the various embodiments of the terminal 10 can include, but are not limited to, mobile terminals/stations, cellular telephones, personal digital assistants (PDAs) having wireless communication capabilities, portable computers (e.g., laptops) having wireless communication capabilities, image capture devices such as digital cameras having wireless communication capabilities, gaming devices having wireless communication capabilities, music storage and playback appliances having wireless communication capabilities, as well as portable units or terminals that incorporate combinations of such functions and sensor networks.
The MEM 10B may be of any type suitable to the local technical environment and may be implemented using any suitable data storage technology, such as semiconductor-based memory devices, magnetic memory devices and systems, optical memory devices and systems, fixed memory and removable memory. The DP 10A may be of any type suitable to the local technical environment, and may include one or more of general purpose computers, special purpose computers (e.g., the ASIC/FPGA 10E), microprocessors, digital signal processors (DSPs) and processors based on a multi-core processor architecture, as non-limiting examples.
Now are detailed certain aspects of the invention that were generally stated above. Recall that in an embodiment of the invention the spectrum sensing operation can be divided into three main portions. First, a coarse wideband spectrum sensing, then narrowing the band and doing a finer band power analysis. Finally the selected candidate band is fed to a spectrum detection algorithm and carefully analyzed. In addition to these a low power envelope detector is constantly in the background detecting signals which emerge at long intervals and then give information that triggers spectrum sensing when necessary to detect these infrequently appearing signals. The spectrum sensing steps are presented in more detail below, but recall that the low power envelope detection operation may be continually operating throughout all or only some of the three stages detailed below. The radio frequency RF envelope detector can be used as controlling the RF gains of the first and second stages (both coarse and fine spectrum sensing branches) to reduce the dynamic ranges of A/D conversions. This information can be also used to aid the spectrum hole search process.
The first stage is coarse sensing over a wideband spectrum. Generically this is considered as searching a wideband spectrum with low resolution to identify bandwidth slices having low or no signal levels. Wideband coarse spectrum sensing can be done with a high speed and low resolution analog to digital converter ADC combined with digital signal analysis (using a fast Fourier transform FFT or equivalently a discrete Fouriertransform DFT, for example). This gives a snapshot of the complete spectrum under study, and the low resolution conserves battery power in the mobile device. Higher resolution over a wide bandwidth would be quite power intensive.
The term wideband spectrum does not imply (nor does it exclude) the whole frequency band over which the cognitive radio CR is capable of receiving, but is a relative term in view of other portions of that wideband spectrum that are identified, selected and decided below in the other stages. It may in fact be the entire frequency band over which the CR is capable of receiving or it may be something less. For example, the scope of the wideband can be decided in co-operation with other CRs (e.g., limiting the wideband to only those frequencies both CRs can receive where one CR has lesser frequency capabilities than the other), it can be decided autonomously by the individual CR based on knowledge of signal activity in the area in which it operates (e.g., it excludes certain bandwidths due to its past experience or information it received that those bandwidths in this particular geographic area are generally not productive for finding free spectrum, such as the UTRAN frequencies in a city).
Similarly, The terms higher and lower resolution are relative to one another rather than absolute. Resolution in the wideband search gives a certain granularity, and resolution in the narrower band search gives a different granularity. So long as the granularity of the wideband search is less than that of the narrower band search, the wideband search is of lower resolution.
Still within the low resolution wideband search, the coarse power measurement may be enhanced in an embodiment according to frequency bins. A frequency bin simply refers to a slice of the wideband spectrum that is determined from the sampling frequency fs at the ADC. For example, a first frequency bin may be defined from 0 Hz to fs, a second frequency bin defined from fs to 2*fs, a third frequency bin may be defined from 2*fs to 3*fs, and so on for the entire wideband spectrum. Note that processing by frequency bins is simply a convenient way to parallelize the wideband processing; the total of all the frequency bins spans the wideband spectrum and so there is no narrowing of the wideband or of individual bins yet. Processing by frequency bins is important because of the folding of spectrum due to frequency sampling at the ADC. The analysis of the frequency bins to find those with the least or no signal level (least again being relative to other frequency bins) can be implemented for example with a FFT block. Note that since the frequency bins span the wideband spectrum, this FFT per frequency bin is simply an implementation of the low resolution wideband spectrum searching that is more broadly stated above. However, it is noted that in certain cases (e.g., with a sufficiently high ADC sampling speed) the frequency bin processing may be dispensed with in favor of serially sampling the wideband spectrum.
At this stage of the signal processing there are what is termed bandwidth slices (to avoid confusion with the term frequency bin though the two imply the same concept) of the wideband spectrum, and for each bandwidth slice it is known with low granularity the signal levels within that slice. Signal level may be defined in different ways according to what is deemed most successful at finding the spectrum holes; it may be volume of signal activity, it may be signal strength, but most appropriately it is a combination; signals with higher strength are weighted according to that strength so as to minimize the influence of Gaussian noise signals on the spectrum sensing process. The weighted signals within a bandwidth slice are summed to get a net result. This net result is then the ‘signal level’ for that bandwidth slice. In this coarse wideband processing stage, the extent of the wideband is limited by eliminating some of those bandwidth slices that have the higher signal level, since the goal is to find free spectrum. So entering the next stage we have only those bandwidth slices that are identified as having low or no signal levels and we do not expend further processing or battery power on further examining the slices with higher signal levels.
In the second stage of the spectrum sensing processing is a gradual search of those identified bandwidth slices with no or low signal levels. More generally, the identified bandwidth slices are searched with a higher resolution than was used to coarse-search the wideband spectrum, and candidate frequency slices are selected as those bandwidth slices having the least signal level after the higher resolution searching. This higher resolution searching can be done in a number of ways, including (but not limited to):
All of these searching techniques can be implemented by configuring the specific blocks shown at
Further in the second stage the most promising candidate frequency slices are selected. Just as in the first coarse-sensing stage, the overall bandwidth under consideration at the close of the second stage is reduced by eliminating those slices with higher signal level. The candidate frequency slices are selected and ranked in increasing order of signal level from lowest (or zero) to highest. These candidate frequency slices are then reduced in number as compared to the number of selected bandwidth slices that began the higher resolution processing in the second stage, regardless of whether or not the range per slice is narrowed during the second stage processing. When narrowed sufficiently it may be convenient to term these candidate frequency slices as candidate channels.
At the close of the second processing stage then is a ranked order of the candidate frequency slices (channels) from the lowest signal level to the highest. The third processing stage is simple; execute a spectrum detection algorithm on the slices in the order of the priority ranking until one is found that is decided to have sufficiently free spectrum. There are several spectrum detection algorithms known in the art and more sure to be developed as the cognitive radio concept becomes closer to practical deployment, and these teachings are not limited to any particular one whether implemented in software, hardware, firmware, or some combination. One well known algorithm is cyclostationary feature detection, for example. This is a fairly power intensive function, and so for the handheld terminal application the rank order is important; parallel processing of multiple candidate slices is seen to be too power inefficient. Once a decision is made from the spectrum detection stage that one of the candidate frequency slices is suitable (i.e., it has enough free spectrum), then the cognitive terminal transmits in the decided candidate frequency slice.
Exactly where the cutoff is to decide what is sufficiently free spectrum is for the CR to decide. At early stages of cognitive radio development the threshold might be much higher than at later stages when the general principles are set and more advances are made to fine-tuning which would enable the CRs to exploit ever smaller portions of free spectrum. However defines, it is convenient that the CR has stored in its local memory a threshold for what is considered enough free spectrum. Then the decision is based on comparing the stored threshold to the output of the third stage spectrum detection algorithm for each candidate slice in turn, another low-power operation. The stored threshold may be a fixed value or may be a function of different parameters (e.g., size of candidate slice, remaining battery power in the CR, etc.).
Now that the details of the various embodiments are laid out,
At the coarse power detector 304 the signal is amplified again at a variable gain amplifier 316 and downconverted to baseband at a mixer 318 that uses the output of a phase locked loop 322 driven by a voltage controlled oscillator 320. The coarse power detector 304 mixes the wideband-signal to baseband, filters it and performs a FFT to it. Most of the blocks of the coarse stage 304 are tunable in some properties though not specifically marked on the figure to prevent obscuring the more pertinent features. Sampling rate at the VCO 320 is controlled by a controller 310 to match the relevant frequency. The baseband signal is converted to digital at an ADC 324, passes through a low pass filter 326, the samples are then decimated by decimator 228 to a convenient number per interval, and then converted to a scaler 330. The scaler passes the samples through a FFT block where the signal is converted from time domain to frequency domain and at the FTT analyzer/power detector 336 a low-resolution analysis is made of the signal level. This is repeated among the two or more branches through the coarse power detector 304 for the various bandwidth slices, and the controller 310 or 310′ drops those slices with higher signal level and retains those slices having the lower (or zero) signal levels for processing in the fine power detector 306. Other techniques apart from FFT/DFT are known to detect frequency spectrum in baseband processing, such as parallel filter banks. These and other substitutions can be readily made to the above exemplary embodiment of baseband processing, and without departing from these teachings the overall radiofrequency architecture can be direct conversion, low-intermediate frequency, heterodyne, or any of several other RF architectures known in the art.
After the mixer 318 the path is divided between I and Q branches. The main controlling is done to signal levels by tuning the VGA 316 and the scaler 330. The information for the tuning is collected from the ADC 324, power meter 234 and the FFT 332. Since the information for the most promising bands is conveyed to the fine power detector 306, a control from the FFT Analyzer 336 is drawn to the VCO 346 of the fine power stage 306 to represent that. The envelope detector 302 is shown as measuring the signal after the mixing 318, but the envelope detector input could be located to another suitable circuit node as well.
In the fine power detector 306 the tunable bandpass filter 312 is tuned to allow passage of a signal within one of the slices identified at the coarse power detector 308. Other identified slices are passed either through other branches of the fine power detector 306 or serially along the same branch. The VCO 346, PLL 348 and mixer 342 are similar to those described with reference to the coarse power detector 304. The variable gain amplifier 350, the tunable bandstop filter 352, ADC 354, and the tunable bandpass filter 356 are configurable for the fine power detection done in this stage 306, and can be dynamically operated to narrow the bandwidth slice by filtering, adjusting the frequency of the signal via the VCO, increasing resolution via the ADC using higher sample rates and/or reduced bandwidth of the slice, tuning the FFT parameters, and/or canceling RF interferers on either side of the signal path ADC 354 divides.
The fine power detector 306, after band-filtering 312 and amplification 314, contains a tunable bandstop filter 340. The bandstop filter is important for reducing the dynamic range requirements of the ADC 354. It can be tuned to filter out for example the largest nearby blocking signal. Several alternatives exist to realize this filter, but the exemplary implementation presented at
In some cases it will be possible to align coarse and fine spectrum sensing processes into different time slots (periods of time). In those cases many of the blocks shown in
It is anticipated that with current technologies, most of the signal processing functionality will be done with dedicated hardware elements and control with software. This is not a limit to the invention however, which is not dependent on a specific hardware/software/firmware implementation.
The above-described spectrum sensing technique is complete for spectrum detection with potentially low power consumption. It is suitable for wideband detection and mobile terminals. This solution can effectively perform RF pre-processing functions that reduce the resolution of A/D conversion and allows the optimization of power consumption with respect to signal levels over the time and bandwidth.
In general, the various embodiments may be implemented in hardware or special purpose circuits, software (computer readable instructions embodied on a computer readable medium), logic or any combination thereof. For example, some aspects such as the sequence generator may be implemented in hardware, while other aspects may be implemented in firmware or software which may be executed by a controller, microprocessor or other computing device, although the invention is not limited thereto. While various aspects of the invention may be illustrated and described as block diagrams, flow charts, or using some other pictorial representation such as
Embodiments of the invention may be practiced in various components such as integrated circuit modules. The design of integrated circuits ICs is by and large a highly automated process. Complex and powerful software tools are available for converting a logic level design into a semiconductor circuit design ready to be etched and formed on a semiconductor substrate.
Programs, such as those provided by Synopsys, Inc. of Mountain View, Calif. and Cadence Design, of San Jose, Calif. automatically route conductors and locate components on a semiconductor chip using well established rules of design as well as libraries of pre-stored design modules. Once the design for a semiconductor circuit has been completed, the resultantdesign, in a standardized electronicformat(e.g., Opus, GDSII, or the like) may be transmitted to a semiconductor fabrication facility or “fab” for fabrication.
Various modifications and adaptations may become apparent to those skilled in the relevant arts in view of the foregoing description, when read in conjunction with the accompanying drawings. However, any and all modifications of the teachings of this invention will still fall within the scope of the non-limiting embodiments of this invention.
Although described in the context of particular embodiments, it will be apparent to those skilled in the art that a number of modifications and various changes to these teachings may occur. Thus, while the invention has been particularly shown and described with respect to one or more embodiments thereof, it will be understood by those skilled in the art that certain modifications or changes may be made therein without departing from the scope and spirit of the invention as set forth above, or from the scope of the ensuing claims.
