Patent Application
20140146857
The field of the invention is that of medium-access control (MAC) in a multi-hop wireless mesh network (also referred to as ad hoc wireless mesh network) comprising a plurality of nodes.
More specifically, the invention pertains to a technique for setting (also referred to as assigning or allocating) frequency channels in such a network.
The present invention can be applied, notably but not exclusively, in the case where each of the nodes comprises or is connected to at least one sensor (e.g. a seismic sensor).
A first known solution, for setting frequency channels in a multi-hop wireless mesh network, is described in U.S. Pat. No. 7,773,457 (Crice et al.), which describes a method for acquiring seismic data using a wireless network comprising a number of individual data acquisition modules (that are configured to collect seismic data and forward these data to a central recording and control system). It is proposed to assign different frequencies to the acquisition modules, so that they don't interfere with one another. In other words, it is proposed a Frequency Division Multiple Access (FDMA), i.e. a channel access method allocating to each acquisition module one or several channels (also referred to as frequency bands). Then different acquisition modules can transmit concurrently, thereby increasing the seismic data read out rate.
A major drawback of this first known solution is that the frequencies have to be assigned to the acquisition modules by a third party (centralized allocation). The acquisition modules cannot autonomously and independently set their frequency.
A second known solution, for setting frequency channels in a multi-hop wireless mesh network, is described in the following article: “A New Multi-Channel MAC protocol with On-Demand Channel Assignment for Multi-Hop Mobile Ad Hoc Networks” by Shih-Lin Wu, Chih-Yu Lin, Yu-Chee Tseng and Jang-Ping Sheu. The authors divide the bandwidth in one control channel and several data channels. The purpose of the control channel is to resolve contention/collision on data channels (medium access issue) and to decide which data channels to be used by which hosts (data channel assignment issue). It is proposed a new multi-channel MAC protocol which can be applied to both FDMA and CDMA technologies. The main idea of this protocol is as follows. For a mobile host A to communicate with a mobile host B, A will send on the control channel a RTS (Request-To-Send) to B carrying its FCL (Free Channel List). Then B will match this FCL with its CUL (Channel Usage List) to identify a data channel (if any) to be used in their subsequent communication and reply a CTS (Clear-To-Send) to A on the control channel. On receiving B's CTS, A will send a RES (reservation) packet on the control channel, to inhibit its neighbourhood from using the same data channel. Similarly, the CTS will inhibit B's neighbourhood from using that data channel. Finally, a data packet will be transmitted from A to B on that data channel.
This second known solution obviates the aforesaid major drawback of the first known solution since there is no need for an assignment of the data channel by a third party (no centralized allocation).
However, this second known solution requires the use of wideband signals instead of narrowband signals (because of radio regulations, as detailed below), which leads to a worse sensitivity, thus to a lower radio link budget for each mobile host, and therefore to a lower battery life (since a higher transmit power is needed).
Indeed, in order to increase the battery life comprised in a node of the multi-hop wireless mesh network, one has to lower the transmit power of this node. To keep the range of the system (i.e. to keep the radio link budget), one has therefore to improve the sensitivity of the system (the radio link budget, noted RLB is given by: RLB=G+P−S, with G the antenna gain, P the transmit power and S the sensitivity). The sensitivity can be improved by using narrowband signals (because the sensitivity depends from the noise power, which itself depends from the signal bandwidth). But because of radio regulations (FCC, CE-ETSI . . . ), FHSS (“Frequency-hopping spread spectrum”) shall be used when using narrow bandwidth signals (bandwidth lower than 500 kHz), e.g. in the unlicensed 2.4 GHz ISM band (“Industrial, Scientific and Medical band”). From these radio regulations, FHSS might be not mandatory but in that case the allowed transmit power would be too much low to keep the range. So one has to use FHSS to get a sufficient power level.
But as the second solution needs one control channel and several data channels, and a mechanism of RTS/CTS on the control channel, this second known solution cannot use narrowband signals because FHSS requires an equal use of the frequencies. Thus, this second known solution cannot use narrowband signals to improve the sensitivity, and thus the battery life (since a higher transmit power is needed in that case).
A particular embodiment of the invention proposes a method for setting frequency channels in a multi-hop wireless mesh network comprising a plurality of nodes. Each of said nodes hops on frequency channels, with a hop period, according to a frequency channels hopping sequence. All data packets transmitted by said nodes have a duration strictly longer than said hop period. When a given node of said plurality of nodes is in a first transmit mode in order to transmit a data packet, it carries out steps of:
The general principle is that of using a frequency channels hopping sequence (i.e. a FHSS technique) to allow any node to select a transmit frequency channel. Since packets have a duration strictly longer than the hop period and thanks to the random behaviour of the transmission event, several transmissions can occur at the same time on different frequency channels. Thus there is no need for a frequency channel assignment by a third party (no centralized allocation). Each node can autonomously and independently set its frequency channel with an increased probability to get simultaneous links on different frequency channels.
This invention uses the FHSS technique in an uncommon way, which leads to an intrinsic implementation of FDMA with an independent and autonomous frequency channel setting by each transmitter node. Thanks to FDMA the throughput of the network is not degraded by collisions and thanks to FHSS the proposed solution is compliant with the radio regulation and allows the use narrowband signals which leads to a better sensitivity, thus to a higher radio link budget and therefore to a higher battery life (since a lower transmit power is needed).
Furthermore, FHSS minimizes the effect of frequency fading and then strengthen the radio link in obstructed environments. Then this allows to implement a multi-hop wireless mesh networks in obstructed environments without any frequency channel planning, in order to ease the installation of these network topologies, especially in the case of wireless seismic networks. Furthermore it leads to an increasing of the battery life of the nodes, through the use of narrow band signals.
According to a particular feature, when a given node of said plurality of nodes is in a receive mode it carries out steps of:
Thus the aforesaid frequency channels hopping sequence is also used to allow any node to select a receive frequency channel.
According to a particular feature, when a given node of said plurality of nodes is in a second transmit mode in order to transmit an acknowledgement packet of a data packet previously received by said given node on a given receive frequency channel, it carries out a step of: transmitting said acknowledgement packet using as transmit frequency channel said given receive frequency channel.
Having the acknowledgement packet (ACK packet) transmitted on the same frequency as the reception frequency allows:
According to a particular feature, when a given node is in the first transmit mode, it carries out a supplemental step of: inserting a random delay before carrying out the transmitting step.
This allows to further increase the probability to get simultaneous links on different frequency channels.
According to a particular feature, said supplemental step of inserting a random delay is carried out before a second attempt to execute the transmitting step, if a first attempt to execute the transmitting step is not successful due to a transmission failure or a collision detection.
According to a particular feature, said random delay is defined by: Trand=Random( )×Thop, where:
Trand is the random delay,
Thop is the hop period, and
Random( ) is a pseudo-random integer.
Thus the random delay is computed easily.
According to a particular feature, said pseudo-random integer Random( ) is drawn from an uniform distribution over an interval [O, CW], with CW a contention window having an integer value.
According to a particular feature, the contention window CW is incremented following a binary exponential way, when the transmitting step can not be executed normally.
According to a particular feature, when a given node is in the first transmit mode, said selecting step comprises:
Thus the selecting step is simple to implement.
According to a particular feature, L is a prime number.
Thus all the frequency channels are used equally.
According to a particular feature, in transmitting step, the data packet is transmitted with a signal having a bandwidth lower than 500 kHz.
As already mentioned above, the use narrowband signals leads to a better sensitivity, thus to a higher radio link budget and therefore to a higher battery life (since a lower transmit power is needed).
In a particular application, each of said nodes comprises or is connected to at least one seismic sensor belonging to the group comprising:
analog or digital seismic sensors having a single vertical component; and
analog or digital seismic sensors having three orthogonal components.
According to a particular feature, said nodes transmit data packets comprising quality control data.
In another embodiment, the invention pertains to a computer program product comprising program code instructions for implementing the above-mentioned method (in any of its different embodiments) when said program is executed on a computer or a processor.
In another embodiment, the invention pertains to a non-transitory computer-readable carrier medium, storing a program which, when executed by a computer or a processor causes the computer or the processor to carry out the above-mentioned method (in any of its different embodiments).
Another particular embodiment of the invention proposes a node belonging to a plurality of nodes comprised in a multi-hop wireless mesh network. Said node comprises:
According to a particular feature, said node comprises the following means, activated when said node is in a receive mode:
Other features and advantages of embodiments of the invention shall appear from the following description, given by way of an indicative and non-exhaustive examples and from the appended drawings, of which:
We first present the general principle of the invention, which relates to a method for setting frequency channels in a multi-hop wireless mesh network comprising a plurality of nodes. For example (see
As illustrated in
As illustrated in
All the data packets 21-24 transmitted by the nodes have a duration which may vary from one data packet to the other, but which is always strictly longer than the hop period Thop. In other words, FHSS is implemented with time slots strictly shorter than the duration of transmitted data packets. This feature (time slots strictly shorter than the minimal data packet duration) allows to increase the probability to get simultaneous links on different frequency channels.
In the embodiment of
In this example, when it is in the transmit mode TX, the node A transmits the data packet 31a using the transmit frequency channel F1, which is given by the frequency channels hopping sequence S. The selected transmit frequency channel F1 is fixed along the entire duration of the data packet 31a. The node B, which is in the receive mode RX, detects the data packet 31a when it is on the frequency channel F1, and then stays on this frequency channel till the end of the data packet 31a. After the data packet 31a has been entirely received, the node B goes into the transmit mode TX in order to transmit the corresponding acknowledgement packet 31b using the same transmit frequency channel F1 (all along the entire duration of the acknowledgement packet 31b). The node A, which is in the receive mode RX, receives the acknowledgement packet 31b, on the frequency channel F1.
Same reasoning applies to the transmission, on the frequency channel F9, of both the data packet 32a (from node A to node B) and the corresponding acknowledgement packet 32b (from node B to node A).
As illustrated in
The node A transmits the data packets 41 and 42 using the transmit frequency channels F1 and F9 respectively, which are each given by the frequency channels hopping sequence S. Each of the selected transmit frequency channels F1 and F9 is fixed along the entire duration of the data packet 41 or 42.
The node C transmits the data packets 43 and 44 using the transmit frequency channels F2 and F7 respectively, which are each given by the frequency channels hopping sequence S. Each of the selected transmit frequency channels F2 and F7 is fixed along the entire duration of the data packet 43 or 44.
In this example, the node 50 is a “seismic sensor unit” comprising:
The node 50 is connected to (or, in an alternative embodiment, integrates) a synchronization interface 58 towards a synchronization source (e.g. GPS, radio or IEEE1588) generating a reference clock. As disclosed below, the microcontroller 54 uses an internal clock (CLK) which is synchronized with the reference clock, using an external signal received by the synchronization interface 58.
The node 50 is connected (e.g. via a string) to at least one seismic sensor 57. In an alternative embodiment, the at least one seismic sensor 57 is integrated in the node 50.
For example, each seismic sensor 57 is:
The read-only memory 56 is a non-transitory computer-readable carrier medium. It stores executable program code instructions, which are executed by the microcontroller 54 in order to enable implementation of the present method (method for setting frequency channels in a multi-hop wireless mesh network), as described above (in relation to
Upon initialization, the aforementioned program code instructions are transferred from the read-only memory 56 to the random access memory 55 so as to be executed by the microcontroller 54. The random access memory 55 likewise includes registers for storing the variables and parameters required for this execution.
All the steps of the present method (method for setting frequency channels in a multi-hop wireless mesh network) can be implemented equally well:
Characteristics of the Node
In a particular embodiment, the node 50 has the following characteristics:
Frequency Channels Hopping Sequence and Hop Set
The RF transceiver 51 hops on 1 MHz spaced channels with a hop period Thop depending on the data rate (see Table 1).
The carrier frequency (also referred to as the reference frequency) F of the transmit frequency channel is derived from:
The sequence of integers is synchronized on an internal clock (CLK) which increments by one every Thop. The internal clock can be synchronized by an external signal (GPS, etc.).
The carrier frequency F is computed as follows:
F(CLK)=F0+ΔF*Sequence(CLK modulo L) (1)
with Sequence(x) a function giving an integer having rank x in the sequence of integers (which itself defines the frequency channels hopping sequence).
Different hop sets can be configured by modifying the base frequency F0. Examples of seauences of integers are given in Table 2.
By default, the sequence number is set to Sequence S1. The sequence number can also be configured by the user.
For example, with: F0=2430 MHz, ΔF=1 MHz, L=17 and Sequence S1, we have:
The values F(0) to F(16) form the frequency channels hopping sequence S shown in
Random Delay
As already mentioned above, in a particular embodiment, a random delay Trand is inserted before each data packet transmission (not before the acknowledgement packets if any).
This random delay Trand is defined as follows:
T
rand=Random( )×Thop,
where:
For example, with CW=7 at 1.2 kbps, Trand is comprised between 0 and 910 ms (7×Thop—1.2 kbps).
The Contention Window (CW) is incremented following a binary exponential way when the transmitter node can not execute normally the transmitting step (because the transmitter node finds the medium busy after a carrier-sense mechanism or when the transmission fails for any reason, as in IEEE 802.11 standard).
Transmit Frequency Channel and Maximal Packet Duration
Data packets are transmitted over several hop periods and they overlap at least two hop periods.
The transmit frequency channel of the data packet is fixed for the entire data packet duration (see
The transmit frequency channel of the data packet is derived from the internal clock value (CLK) at the start of the data packet, according to above equation (1).
Table 3 gives the maximal duration of the data packets according to the data rate.
In a step 61, the node selects a transmit frequency channel as a function of the frequency channels hopping sequence S. In other words, the node selects a transmit time to which is associated a transmit frequency channel of the frequency channels hopping sequence S. In a particular embodiment already discussed above, the selecting step 61 comprises: obtaining a current value of the clock CLK when the selecting step 61 starts; and computing a reference frequency F(CLK) of the transmit frequency channel according to above equation (1).
In a step 62, the node inserts a random delay Trand before carrying out a transmitting step 63.
In the transmitting step 63, the node transmits a data packet using, for the entire duration of this data packet, the selected transmit frequency channel.
In a step 71, the node detects a data packet on a given receive frequency channel of the frequency channels hopping sequence S.
In a step 72, the node checks if it is the destination of the detected data packet.
If the node is the destination of the detected data packet, it goes to step 73, in which the node stays on the given receive frequency channel till the end of the detected data packet.
If the node is not the destination of the detected data packet, it goes to step 74, in which the node hops on the next frequency channel of the frequency channels hopping sequence S, at the end of the current hop period Thop.
An embodiment of the present disclosure provides a technique for setting frequency channels in a multi-hop wireless mesh network, this technique allowing:
An embodiment provides a technique of this kind, allowing to minimize the effect of frequency fading and then strengthen the radio link in obstructed environments.
An embodiment provides a technique of this kind that is simple to implement and costs little.
Although the present disclosure has been described with reference to one or more examples, workers skilled in the art will recognize that changes may be made in form and detail without departing from the scope of the disclosure and/or the appended claims.
| Number | Date | Country | Kind |
|---|---|---|---|
| 12306475.0 | Nov 2012 | EP | regional |
