FR3104373A1 - Method and device implementing this method for accessing a shared channel to respond to a relay request request - Google Patents

Abstract

Title of the invention: Method and device implementing this method for accessing a shared channel to respond to a relay request request This method for accessing a shared channel to respond to a relay request request comprises steps of: - receiving the request from a source device; obtaining a size of a contention window during which the source device is able to receive application messages in response to the request; - obtaining a threshold value and a random waiting time defining a response interval in the contention window, the waiting time is obtained by a mathematical law configured to produce a shorter waiting time for a power received from the higher request; - sending in the channel, in response to the request and in the response interval, of an application message indicating that the candidate device can constitute a relay device, the application message being sent only if the power received from the request is greater than or equal to the threshold value. Figure for the abstract: Fig. 3

Description

Method and device implementing this method for accessing a shared channel to respond to a relay request request

The invention relates to the general field of communications in a wireless access network. It relates more particularly to a mechanism for relaying data in a cellular network. It can be applied in a privileged but non-limiting way in a D2D device-to-device (for “Device to Device”) communication network.

In the current context of continuous growth in the number of connected objects, communication networks have emerged to support connected objects, such as cellular networks of the LTE-M type (for "Long Term Evolution - Machine" in English) or NB-IoT type (for "Narrow Band - Internet of Things" in English) or LoRa-WAN type (for "Long Range Wide-Area Network" in English). These LTE-M, NB-IoT and LoRa-WAN protocols are an integral part of the New Radio (NR) standardization process for 5G cellular networks.

A connected object may be in unfavorable conditions to be able to connect to a base station or to benefit from good connection quality, for example when this connected object is far from the base station or when it is located in a environment comprising obstacles (environment known by the English name "deep indoor"), such as a basement. When the connected object perceives a poor connection quality with the base station, it increases the transmission power of these packets, and may have to resend packets not received by the base station. D2D relaying (specified in particular in document TR36.746 Study on further enhancements to LTE Device to Device (D2D), UE to network relays for Internet of Things (IoT) and wearables (release 14) v15.1.1) solves this problem: a network device located close to the connected object and benefiting from a better quality of connection to the base station, plays the role of a relay device between the connected object and the base station.

We designate here by "source device", a device, for example a connected object whose data it sends to a base station are relayed by a relay device.

According to the state of the art, the relay device can be a dedicated network device for relaying, for example a server, or a user device, for example a terminal, for example a telephone.

We remind you that the technique of relaying is different from that of connection sharing. Indeed, when a first terminal shares its connection to the network with a second terminal, the two terminals use the network connection flat rate of the first terminal. Seen from the network, the messages generated by the two terminals are considered to be all messages generated by the first terminal, a network device (the base station for example) generally does not distinguish between the messages generated by each of the terminals . The first terminal allows the second terminal to benefit from its subscription to the network.

On the contrary, a relay device identifies, with the base station, the messages relayed by an identifier of the source device which generated them. The base station or the network can distinguish between messages that the relay device generates by itself, and messages generated by the source device and relayed by the relay device. Unlike tethering, the source device generally does not take advantage of the relay device's subscription. The source device must be subscribed to the network and identified with the base station, the relayed messages are counted on the subscription of the source device and not that of the relay device. The network operator can nevertheless offer financial or commercial advantages to the devices which support the technique of relaying and which apply to ensure the relaying.

It is conceivable that the selection of a relay device for a source device is carried out in a centralized manner by the base station, or in a distributed manner by the source device itself. The present invention is in the context of distributed selection. The article by C.V. Anamuro et al., “Energy-efficient discovery process for mMTC applications”, Wireless and Mobile Networking Conference, Paris, September 2019 (hereinafter referred to as “Anamuro paper”), and the article by A. Awang et al., “RSSI-based forwarding for multihop wireless sensor networks”, Eunice workshop EUNICE9, pp. 138-147, 2009 (hereinafter referred to as “Awang paper”), relate to the distributed selection of a relay device.

According to the state of the art, D2D relaying has three main phases for the source device:
- establishment of a connection and synchronization with the base station;
- Following detection of poor connection quality, discovery (search for possible relay devices) and selection of a relay device; And
- the transmission of data, intended for the base station, to the selected relay device.

FIG. 1 is a timing diagram which illustrates the phase of discovery and selection of a relay device, according to a method of the state of the art described in the Anamuro document cited above. A source device MTD (Machine-Type Device) and user devices UE-1 to UE-5 are connected and synchronized with a base station of a cellular access network. The source device MTD broadcasts, in an initial time interval, a relay request request RR (RR for "Request for Relay" in English). In this request RR, the source device MTD specifies a size W of a contention window during which it is able to receive application messages in response to its request. The devices UE-1 to UE-5 supporting D2D relaying apply when they receive the request RR to play the role of relay device. Each of these candidate devices UE-1 to UE-5 draws a random number between 1 and W, corresponding to a number s of a time slot in the contention window. Each candidate device can respond to the RR request, by sending an RC (Relay Candidate) application message, in the interval corresponding to the selected random number. In the example illustrated by FIG. 1, the candidate devices UE-1 to UE-5 respectively obtain the numbers of the intervals "s": 4, 2, 6, 2, and 3. According to the state of the art, as soon as the source device MTD receives an application message, it selects the device which sent its application message to constitute its relay device.

In the example illustrated by FIG. 1, the source device MTD receives in the interval number 2 two candidacy messages coming from the candidate devices UE-2 and UE-4, these messages collide and are not correctly received by source device MTD. In interval number 3, the source device MTD receives an application message RC from the candidate device UE-5. The source device MTD selects the device UE-5 to be its relay device, and broadcasts a return message F (for "feedback"), this message F indicating that the source device MTD has selected a relay device. Receiving the return message F, the candidate devices UE-1 and UE-3 no longer send the planned candidacy messages in intervals number 4 and 6 respectively.

We note that only the source device can use the channel at the beginning of an interval, to avoid a collision between the return message F and a possible application message RC. Before sending an application message, each candidate device listens to the channel and only sends its message if the channel is free.

Once the source device MTD has selected its relay device UE-5, it sends it the data intended for the base station. The UE-5 relay device acknowledges the reception of this data and relays it to the base station. The UE-5 relay device can be configured to also relay data in the downstream direction, from the base station to the MTD device.

This relay device discovery and selection phase is a phase that may require significant energy consumption in the cellular access network, in particular for the source device.

The Awang document cited above describes a relay device selection method in which the candidate device which has the best connection performance with the base station is selected to be the relay device. The same authors propose in their article "Toward an analysis of energy consumption in multihop wireless sensor networks", International Conference on Intellignet and Advanced Systems, Kuala Lumpur, 2010, to add as a selection criterion the residual energy of the device battery candidate to be a relay. In addition, the selection method described in the aforementioned document favors the selection of the candidate device closest to the base station, thus making it possible to reduce the energy consumption of the relay device and to benefit from good quality of the channel between the relay device and the base station.

Thus the selection methods described in the Awang documents aim to reduce the energy consumption of the relay device, or of the assembly consisting of the relay device and the source device. However, we note that when the relay device is a user device, of the smartphone type for example, the energy consumption of the relay device due to relaying may be negligible compared to its own consumption, especially when the relay device also transmits its own data to the base station. The methods of the prior art exposed above are not optimized to reduce the energy consumption of the source device.

There is therefore a need for a relaying method which limits the consumption of the source device.

The invention relates to a method for accessing a channel shared between several devices of a cellular access network, called "candidate devices", to respond to a relay request request, the relay device being intended to relay messages in the network between a communicating device called "source device" and an access point of said network, the source device being synchronized with said access point. This method implemented by a said candidate device comprises steps of:
- reception, from the source device, of the relay request request;
- Obtaining a size of a contention window during which the source device is able to receive application messages in response to the request;
- Obtaining a random waiting time defining a response interval in the contention window;
- sending in the channel, in response to the request and in the response interval, an application message indicating that the candidate device can constitute a relay device;
this method further comprises a step of obtaining a threshold value, the application message is sent only if a power received from the request is greater than or equal to the threshold value;
the waiting time is obtained by a mathematical law configured to assign a shorter waiting time for said higher received power.

Correlatively, the invention aims at a device of a cellular access network, called "candidate device", able to communicate via a channel shared between several devices of the network to respond to a relay request request, the relay device being intended to relaying messages in the network between a communicating device called "source device" and an access point of said network, the source device being synchronized with said access point.
The candidate device according to the invention comprises:
- a communication module configured to receive, from the source device, the relay request request;
- a module for obtaining a size of a contention window during which the source device is able to receive application messages in response to the request;
- a module for obtaining a random waiting time defining a response interval in the contention window;
- the communication module being configured to send in the channel, in response to the request and in the response interval, an application message indicating that the candidate device can constitute a relay device;
the candidate device further comprises a module for obtaining a threshold value, the communication module being configured to send the candidate message only if a power received from the request is greater than or equal to the threshold value;
the waiting time obtaining module is configured to obtain the waiting time by a mathematical law configured to assign a shorter waiting time for a higher received power.

The characteristics and advantages of the access method according to the invention presented below apply in the same way to the candidate device according to the invention, and vice versa.

In one embodiment of the invention, the candidate device is a relay-supporting user device capable of operating as a relay device. The candidate device according to the invention can be a smartphone type terminal, or a tablet equipped with a subscription card to the cellular network, such as a SIM card (for "Subscriber Identity Module" in English or module of identity of the subscriber) or eSIM (for “embedded SIM” in English or integrated SIM), or any other communicating device capable of connecting to the cellular network and relaying messages from another device to the cellular network access point .

The source device according to the invention can be a connected device or object supporting D2D relaying. The source device can connect directly to the access point of the cellular network, or via the relay device, for example in the event of a poor quality of the direct channel between the source device and the access point, or in the event of impossibility of direct connection to the access point. The source device can be a terminal, for example of the smartphone, computer, tablet, watch, smart home device type, or any other type of connected object.

In accordance with the invention, the access point to the cellular network can be of the eNodeB, advanced eNodeB or gNodeB base station type, also denoted gNb. The access point manages radio cells using different frequency bands on different geographical areas, covered by the cellular access network.

Unlike the methods described in the prior art documents presented above, the present invention favors the devices closest to the source device to give them more chances of being selected as a relay device. The invention therefore makes it possible to reduce the energy consumption of the source device. Since the source device is close to the candidate device according to the invention, it does not require high power to transmit its messages to be relayed, nor repeated sending of these messages.

The invention makes it possible to reduce the energy consumption of the source device during the discovery phase and selection of the relay device, but also during the message transmission phase. Indeed, the closer the relay device, the faster the messages to be relayed will be transmitted and the less the source device will consume energy in the message transmission phase.

The source device can adapt its rate to the maximum possible during the data transmission phase.

In one embodiment, the invention takes into consideration the characteristics of the channel connecting the source device to the candidate device, in particular the propagation loss PL (for “Path Loss” in English).

The comparison between the power received from the request and the threshold value is intended so that only the candidate devices to which the source device is connected via channels with low PLs can compete for the selection of a relay device. Thus, only candidate devices close to the source device can send candidate messages.

We designate here by the candidate device "closest" to the source device, the candidate device whose connection channel with the source device is of better quality, presenting the lowest PL loss. According to this definition, the The closest candidate device may be the one closest in distance to the source device, or a device whose connection to the source device is unaffected by obstacles.

In accordance with the invention, the threshold value obtained can directly be a power. Alternatively, this power can be deduced from an obtained distance, for example a radius of a discovery area in which candidate devices can receive the request from the source device. It can also be deduced from a propagation loss PL.

In one embodiment, the mathematical law according to the invention is a truncated geometric law. The probability that a number s of a response interval (in the contention window of size W) is chosen by a candidate device is defined by:
[Math 1]
where r denotes the distance between the source device and the candidate device obtained from the received power of the request, Rd denotes a distance corresponding to the threshold value, and the parameter b is a constant between 0 and 1 excluded.

The closer the candidate device is to the source device, the shorter the wait time. This law gives a high probability to the closest candidate devices to respond in the first response intervals, and therefore to have a higher chance of selection.

The probability that the relay device is close to the source device is higher when the value of the parameter b is smaller.

In one embodiment, the relay request request includes the parameter b of the truncated geometric law.

In this mode, the source device chooses the parameter b and broadcasts it to all the candidate devices. All these candidate devices obtain their waiting times according to the same parameter b, the candidate device having the best connection channel with the source device having a higher probability to answer the request first and therefore more likely to be selected as a relay device.

This mode allows each source device to specify the parameter b of its choice, and therefore to adjust a compromise between the probability of discovering potential candidate devices on the one hand, and the probability of selecting the candidate device closest to somewhere else.

In another embodiment, the parameter b of the truncated law is obtained from the access point. In this mode, the truncated law is uniform for all the source devices of the cellular access network.

In one embodiment of the invention, and according to a numerical simulation and an analytical verification, an optimized value of the parameter b is equal to 0.4 when:
- the density per m ² of candidate devices in the network is between 0.8×10 ^-4 and 1.2×10 ^-4 inclusive;
- the distance between the source device and the access point is between 500 and 1000 meters;
- the threshold value corresponds to a distance of 300 meters;
- the size of the contention window is equal to 16; And
- a size of the messages to be relayed is equal to 200 bytes.

In one embodiment of the invention, and according to a numerical simulation and an analytical verification, an optimized value of the parameter b is equal to 0.3 when:
- the density per m ² of candidate devices in the network is greater than or equal to 1.3×10 ^-4 ;
- the distance between the source device and the access point is greater than or equal to 750 meters;
- the threshold value corresponds to a distance of 300 meters;
- the size of the contention window is equal to 16; And
- a size of the messages to be relayed is equal to 200 bytes.

The truncated geometric law is only one example of a law that can be used in the invention. Other mathematical laws making it possible to obtain a shorter waiting time for a higher received power can be considered.

In one embodiment of the invention, at least one parameter among the size of the contention window and the threshold value is included in the relay request request. These parameters can therefore be determined and specified by the source device.

In another embodiment, at least one parameter among the size of the contention window and the threshold value is received from the access point.

It is conceivable that the contention window size and the threshold value are obtained in different ways, for example one from the relay request request and the other from the access point.

In one embodiment, the access method according to the invention further comprises a step of obtaining an identifier of the source device and a step of sending this identifier to the access point with the relayed messages.

The candidate device according to the invention can receive the identifier of the source device from the relay request request. In this case, following receipt of the request, the candidate device saves the identifier. If selected to play the role of a relay, it inserts the identifier in a field of the messages that it relays to the access point, so that the access point can distinguish the messages generated by the device source of those generated by the relay device.

Alternatively, the candidate device may only receive the identifier of the source device when it is selected to be a relay device.

The access method according to the invention is part of a set of steps allowing communication in the cellular access network, between the source device and the access point via a relay device. The access method according to the invention is preceded by a step of broadcasting the relay request request by the source device, and it is possibly succeeded by a selection step, by the source device, according to the application messages that it receives, from a relay device among the candidate devices having responded to its request. The source device then broadcasts a return message with an indication of the selected relay device.

The selected relay device can be instructed to relay messages uplink from the source device to the access point and/or downlink from the access point to the source device.

The invention also relates to a system comprising a source device and at least one candidate device in accordance with the invention.

The invention also relates to a computer program on a recording medium, this program being capable of being implemented in a computer or in a candidate device in accordance with the invention, this program comprising instructions suitable for implementation of an access method as described above.

This program may use any programming language, and be in the form of machine code, source code, object code, or intermediate code between source code and object code, such as in partially compiled form, or in any other desirable shape.

In particular, this program can be executed by a microcontroller µC (“micro-controller” in English).

The invention also relates to information or recording media readable by a computer, and comprising instructions of the computer program as mentioned above.

Information or recording media can be any entity or device capable of storing programs. For example, the media may comprise a storage medium, such as a ROM, for example a CD ROM or a microelectronic circuit ROM, or even a magnetic recording medium, for example a floppy disk or a disk. hard, or flash memory.

On the other hand, the information or recording media can be transmissible media such as an electrical or optical signal, which can be routed via an electrical or optical cable, by radio link, by wireless optical link or by other ways.

The program according to the invention can in particular be downloaded from an Internet-type network.

Alternatively, each information or recording medium may be an integrated circuit in which the program is incorporated, the circuit being adapted to execute or to be used in the execution of the access method according to the invention.

Other characteristics and advantages of the present invention will emerge from the description given below, with reference to the appended drawings which illustrate an exemplary embodiment thereof which is devoid of any limiting character. In the figures:
FIG. 1, already described, is a flowchart representing the steps of a discovery and selection phase of a prior art relay device;
FIG. 2 illustrates an architecture of a cellular access network in which an access method according to the invention is implemented according to an embodiment of the invention;
FIG. 3 is a flowchart representing the steps of an access method according to one embodiment of the invention;
FIG. 4 illustrates selection probabilities of the waiting time of candidate devices in accordance with one embodiment of the invention to constitute relay devices;
FIG. 5 illustrates an evolution of the probability of discovery of a candidate device as a function of the density of candidate devices in a discovery zone, according to an exemplary embodiment of the invention;
FIG. 6 illustrates an evolution of the average number of response intervals used in the contention process as a function of the density of candidate devices in a discovery zone, according to an exemplary embodiment of the invention;
FIG. 7 illustrates a PDF function (Probability Density Function) of the distance between a source device and a relay device, according to an exemplary embodiment of the invention;
FIG. 8 illustrates the impact of a parameter b of a truncated geometric function on the probability of discovery of candidate devices, according to an example embodiment of the invention;
FIG. 9 illustrates the impact of a parameter b of a truncated geometric function on the average number of intervals used in the contention process, according to an example embodiment of the invention;
FIG. 10 illustrates the impact of a parameter b of a truncated geometric function on the PDF function of the distance between the source and relay devices, according to an example embodiment of the invention;
FIG. 11 illustrates the evolution of the total energy consumption of a source device as a function of the density of the candidate devices present in a discovery zone, according to an example embodiment of the invention;
FIG. 12 illustrates the evolution of the optimal value of parameter b as a function of the density of candidate devices in the discovery zone, according to an exemplary embodiment of the invention;
FIG. 13 illustrates a functional architecture of a candidate device, in accordance with the invention, according to one embodiment;
Figure 14 illustrates the hardware architecture of a candidate device according to one embodiment of the invention.
Detailed Description of Embodiments

FIG. 2 illustrates an architecture of a cellular access network NET in which an access method in accordance with the invention is implemented according to one embodiment of the invention.

In this embodiment, the NET network is of the 4G (4th generation) type. The network NET includes an access point BS to this network, which is a base station, for example of the advanced eNodeB type. Several terminals UE (for "User Equipment" in English) and connected objects MTD (for "Machine-Type Device" in English) are connected to the network, under the coverage of the access point BS. Some MTD devices are directly connected to the BS access point. Other devices MTD connect to the access point BS via relay devices formed by terminals UE which support D2D type relaying.

In the mode described here, the terminal devices UE and the connected object devices MTD form two homogeneous independent Poissonnière distribution systems in the network NET, of the H-PPP type (for “Homogenous Poisson Point Process”). The respective densities of the UE devices and of the MTD devices are denoted by λu and λm.

We assume that the BS access point allocates frequencies specific to D2D relaying (between MTD devices and UE relay devices), and which do not interfere with the frequencies used for direct cellular communication between devices in the NET network ( UE and even MTD) and the access point BS.

FIG. 3 is a flowchart representing the different steps of the access method, implemented by a UE device, in accordance with one embodiment of the invention. The device UE implementing the access method in accordance with the invention is capable of being selected by a device of the connected object type MTD, called the source device. If the candidate device UE is selected, it forms a relay device for the source device MTD and relays data between this source device MTD and the access point BS in the uplink direction and/or in the downlink direction.

The flowchart of FIG. 3 represents steps E10 to E60 of the access method according to the invention, and steps F10 to F70 of a method according to the state of the art, for discovery and selection of a device relay, implemented by a source device MTD.

During a step F10, the source device MTD connects and synchronizes with the access point BS. The source device MTD measures the propagation loss Lm,b between itself and the access point BS.

We assume here that this propagation loss exceeds a certain threshold below which the source device MTD determines that its connection with the access point BS is of poor quality. The source device MTD broadcasts, during a step F20, a packet comprising a relay request request RR on a channel dedicated to the synchronization of D2D relaying, this channel being shared between several devices of the cellular access network NET.

In the mode described here, the request RR includes an identifier ID of the source device MTD.

During a step F30, the source device MTD switches
from a sender state, defined on the shared channel, to a receiver state and waits for a response to its RR request.

During a step E10 of the access method according to the invention, the candidate device UE receives the request RR and stores in its memory the identifier ID of the source device MTD.

During a step E20, the candidate device UE obtains, from a power received from the request RR, a propagation loss Lm,u on the connection channel between itself and the source device MTD. To measure Lm,u, the candidate device UE compares the power of reception of the request RR with the power of transmission of the latter by the source device MTD, the power of transmission possibly being defined in the request RR or known in the network from configuration data obtained from the access point BS.

During step E20, the candidate device UE obtains a decisive threshold value Lth for the candidate device to respond or not to the request RR. The threshold value Lth can be included in the request RR or be obtained from configuration data received from the access point BS. In the mode described here, the threshold value Lth is a maximum propagation loss value that must not be exceeded in order to be able to respond to the request RR. This value Lth corresponds to a threshold distance Rd defining a radius of a candidate device discovery zone centered on the source device MTD.

Only the candidate devices situated in this discovery zone of radius Rd can thus respond to the relay request request RR.

The candidate device UE compares during step E20 the propagation loss Lm,u with the threshold propagation loss Lth.

In the mode described here, the threshold value corresponds to a path loss. Alternatively, the threshold value may correspond to a minimum threshold power of reception of the request RR. The candidate device UE responds to the request RR only if the power received is greater than or equal to the threshold power. Alternatively, the threshold value can correspond to a distance (for example in meters). The candidate device responds to the request RR only if the distance which separates it from the source device MTD is less than or equal to the threshold distance.

We assume here that the propagation loss Lm,u on the connection channel between the source devices MTD and candidate UE is lower than the threshold value Lth. The candidate device UE is therefore eligible to respond to the request RR and possibly be selected to play the role of a relay device.

During a step E25, the candidate device UE obtains a size W of a contention window during which the source device MTD is able to receive candidate messages in response to its request RR. The size W can be included in the request RR or obtained from configuration information received from the access point BS.

During a step E30, the candidate device obtains a random waiting time defining a response interval in the contention window. A response interval number s is between 1 and size W inclusive. Each number s has a probability ρ of being chosen. The wait time is obtained by a mathematical law configured to assign a shorter wait time for a higher received power of the RR request.

In the mode described here, the mathematical law is a truncated geometric law, the probability that a number s of said response interval in said contention window is chosen by a candidate device being defined by equation (1) already cited.

where r denotes the distance between the source device MTD and the candidate device UE obtained from the received power of the request RR (or the propagation loss Lm,u), Rd denotes the distance corresponding to the threshold value, and the parameter b is a constant between 0 and 1 excluded.

Other mathematical laws favoring the devices closest to the source device MTD can be considered.

During a step E40, the waiting period expires, the candidate device listens to the shared channel, and if the channel is free at the end of the waiting period, the candidate device UE sends to the source device MTD in said channel, in response to the request RR and in the response interval number s, an application message RC indicating that the candidate device UE can constitute a relay device.

The source device MTD receives the application message RC during a step F40. Indeed, the source device MTD checks at each response interval in the contention window whether it receives an application message. If it receives nothing, it increments during a step F50 a counter of the intervals already passed in the contention window and continues to wait (F30) for the reception of a response to its request RR. The first candidate device that transmits an RC message without collision wins the contention process and is selected as the relay device.

We assume here that the selected candidate device UE is the first to send an application message RC, and that this message is received by the source device MTD without collision. The source device MTD selects, during a step F60, the device UE to constitute its relay device, and broadcasts a return packet F in the interval number s+1 to inform the candidate device UE that it is selected, and informing the other candidate devices possibly present in the discovery zone that the contention is over and that it has found and selected its relay. Thus, if other candidate devices had scheduled to respond to the RR request in subsequent response intervals, they cancel this scheduling.

In addition, the source device MTD announces in the return packet F to the selected candidate device UE the coding and/or modulation scheme that it will use for the transmission of messages to be relayed.

The candidate device UE receives the return packet F during a step E50. In this example, the candidate device UE learns that it is selected, it then waits to receive messages to be relayed for the source device MTD.

During a step F65, the source device MTD sends its messages to be relayed to the relay device UE. +The latter receives them during a step E60, acknowledges their receipt, then relays them to the access point BS by inserting the identifier ID of the source device MTD. The source device MTD receives the acknowledgment of receipt of its messages by the relay device during a step F70.

In another example, the source device MTD does not receive an application message and determines during step F50 that the contention window has expired. In this case, the source device MTD has no relay device at its disposal and it sends its messages directly to the access point BS (step F55).

FIG. 4 illustrates PMF (Probability Mass Function) mass functions of probabilities of selection of the waiting time in candidate devices in accordance with the invention to form relay devices, according to an example implementation of the access method according to the invention.

In this example, the source device MTD is located at the origin of a discovery zone of radius Rd = 300 meters, four candidate devices are located in this discovery zone: UE-1, UE-2, UE3, and UE -4 which are respectively located at 30, 120, 210, and 300 meters from the source device MTD. In this example, the value of parameter b is equal to 0.6 and the size W of the contention window is equal to 8 response intervals.

Figure 4 shows that the closest UE-1 candidate device has the highest probability to respond to the first response interval (s=1), while the farthest UE-4 candidate device has the lowest probability. to answer in this first interval (s=1) and the highest probability to answer in the last interval (s=W). Thus, the method of the invention makes it possible to increase the chances that the closest candidate device UE-1 will be selected to constitute the relay device.

An analytical model and simulations have been developed to demonstrate the advantages of the method of the invention.

Analytical model of propagation:

In these simulations, we assumed that all devices in the network have a fixed transmit power Ptx. We model the signal propagation considering a normal logarithmic masking model. Thus, the received power Prx is defined by:
[Math 2]
Or is the Euclidean distance between the transmitter and the receiver, K and α are the propagation loss factor and exponent respectively, and χ is a Gaussian random variable with a variance σ ² . We assume that χ is constant during the discovery phase. We also consider that each communication mode (direct cellular and D2D) has its own K and α values.

To simplify the analysis, we use the displacement theorem (HS Dhillon and JG Andrews, “Downlink Rate Distribution in Heterogeneous Cellular Networks under Generalized Cell Selection,” in IEEE Wireless Communications Letters, vol. 3, no. 1, pp. 42 -45, February 2014, lemma 1). Thus, the initial homogeneous H-PPP Poisson distribution model of the UE candidate devices is replaced by an H-PPP model with a density without masking, but a modified distance r taking into account the masking:
[Math 3]
[Math 4]
Equation (2) can then be rewritten: [Math 5]

Equation (5) is a distance-based path loss model. This model can be used during step E20 by the candidate devices UE. In this document, the analysis of the performance of D2D communications (MTD-UE channel) is based on the model of equation 5 which makes it possible to obtain a PLR propagation loss rate by considering the distance and the masking effects. due to the obstacles between source MTD and candidate UE devices. there is only one base station and therefore there is no need to use the displacement theorem.

Analytical model of energy consumption

In order to calculate the total energy consumption E _{m, total} of the source device MTD, we consider a model based on the energy states. In this model, the source device MTD can operate in one of the following states: active (transmitter T or receiver R), or idle I. Each state has constant power requirements:
[Math 6]
where P _m;T , P _m;R and P _m;I represent the power consumption of the source device MTD (power consumption per unit time) at states T, R and I respectively, and t _m;T , t _m;R and t _m;I are the proportions of time spent in each of the states T, R and I.

Analytical model for the discovery area

Considering the modified model of the H-PPP system described by equations (3) and (4), a number N of candidate devices UE located inside a discovery zone of radius Rd, has a Poisson distribution. The probability that the number N of candidate devices is equal to n can be defined by:
[Math 7]

The radius Rd of the discovery zone is a parameter of the discovery phase. The greater this radius, the greater the number of candidate devices present in the discovery zone, and therefore the probability of collision between their RC candidate messages increases. The smaller the radius Rd, the probability of finding a candidate device in the discovery area is low.To avoid an empty discovery area, we determine the minimum threshold value of the distance Rd so that the probability of finding at least one EU candidate device
in the discovery zone is greater than 98%, which is equivalent to a probability of finding no candidate device (P(N=0)) less than or equal to 0.02.

Let us consider an example where the density of the candidate devices λu= 10×10 ^-4 UE per square meter, the masking parameter σ=8dB, and the parameter α=4, the radius Rd allowing to obtain a probability P(N=0 ) <=0.2 is greater than or equal to 286 meters.

Probability of receiving RC application message(s)

In each response interval, numbered s, the candidate devices UE form an independent PPP system. At a given response interval, all candidate devices UE located at a distance r from the source device MTD have the same probability ρ(s;r) of transmitting an application message RC, this probability is defined by equation (1) . Thus, the candidate devices form a non-homogeneous PPP system IPPP (for “Inhomogeneous PPP”) centered around the source device MTD and with a density function λ(r)=ρ(s;r) λ′u.

Using the definition of an IPPP system, we can infer the probability of receiving n number of RC application messages within the time slots:
[Math 8]

where Xs is the intensity measure defined by:
[Math9]

The source device MTD successfully receives an application message RC in the response interval s if only one candidate device UE responds in this interval. The probability that a single application message is received (F40) at interval s is denoted by P ₁ (s)=P(N=1).

Since in each response interval the candidate devices UE form an independent system of the PPP type. Thus, the P _disc relay discovery probability can be obtained as:
[Math 10]

Figure 5 shows the evolution of the probability of discovery (of receiving application messages) as a function of the density of candidate devices UE in the discovery zone for a contention window of size W equal to 8, 16, or 32 intervals, a masking parameter σ=8dB, and a radius Rd of the discovery zone of 300 meters. In the example illustrated by FIG. 5, the candidate devices in accordance with the invention obtain their waiting times by a truncated geometric law whose parameter b is equal to 0.6.

In figure 5, curves obtained by the analytical equation (10) and by numerical simulations are superimposed, which confirms the accuracy of the equation (10). Figure 5 shows that as the density of UE candidate devices increases, the relay discovery probability increases until it reaches a maximum and then decreases due to collisions between the candidate messages. The effects of collisions decrease by increasing the size W of the collision window.

Number of response intervals that occurred in the contention process

Let S be a discrete random variable that represents the number of response intervals used in the contention process. This process ends as soon as an application message RC is received by the source device MTD or when the maximum number of slots W is reached. For S between 1 and W-1, the probability that the contention process terminates in a response interval number s is equivalent to the probability of a reception failure in the first interval until interval number (s-1 ) and successful reception at interval number s. Thus the probability that the number of response intervals used S is equal to s is defined by:
[Math 11]

The number S of intervals used in the contention process is equal to the size W when the first RC application message is received alone at the last interval in the contention window, or when no application message is received, or is received alone in the contention window. The probability P(S = W) is defined by:
[Math 12]

We derive from equations (11) and (12) the average number of response intervals used in the contention process:
[Math 13]

Figure 6 shows the evolution of the average number of response slots used in the contention process as a function of the density of candidate devices UE in the discovery zone, for a size W equal to 8, 16 or 32 slots, a masking parameter σ=8dB, and a radius Rd of the discovery zone of 300 meters. In the example illustrated by FIG. 6, the candidate devices in accordance with the invention obtain their waiting times by a truncated geometric law whose parameter b is equal to 0.6. In FIG. 6, curves obtained by analytical equation (13) and by numerical simulations are superimposed, which confirms the accuracy of equation (13). Figure 6 shows that the lower the density of candidate devices UE, the longer the source device MTD waits to receive a candidate message. On the other hand, when the density of candidate devices UE increases, the number of slots used decreases until it reaches a minimum value, then it increases due to the increase in the number of collisions.

Distribution of the distance between the MTD source device and the selected candidate device

The distance between the source device MTD and the relay device is an important parameter which makes it possible to adapt the modulation and coding scheme MCS (for “Modulation and Coding Schema”) in the transmission phase of messages to be relayed. Let us denote by R the distance between
the source device MTD and the relay device UE, the cumulative distribution function of R can be defined by:
[Math 14]

where the distance r is between 0 and the radius Rd inclusive, P(S = s) is the probability that an RC application message will be successfully received within the response interval s, and P(R<r|s ) is the probability that the selected relay device is located inside a disk of radius r given that an RC application message is successfully received within the response interval s. This probability can be defined by:
[Math 15]

By differentiation from equation (14), the density function PDF of the distance between the source device MTD and the candidate device UE selected as the relay device is defined by:
[Math 16]

Figure 7 illustrates the evolution of the PDF function of the distance between the source device MTD and the relay device, for a density of candidate devices λu = 100 x 10 ^-6 devices per m ² , a radius of the discovery zone Rd = 300, masking dB = 8 dB, and size W = 8, 16 or 32 intervals. In the example illustrated by FIG. 7, the candidate devices in accordance with the invention obtain their waiting times by a truncated geometric law whose parameter b is equal to 0.6.

Impact of parameter b in an embodiment of use of the truncated geometric law

Figures 8, 9, and 10 illustrate the impact of parameter b on the discovery probability, the average number of intervals used in the contention process, and the PDF function of the distance between the source and relay devices, respectively.

In the exemplary implementation of the method of the invention, illustrated by FIGS. 8, 9 and 10, the radius of the discovery zone Rd is fixed at 300 meters and the size W of the contention window is fixed at 16 intervals. The masking parameter is equal to 8 dB. the parameter b takes the values of 0.2, 0.4, 0.6 and 0.8. When the parameter b tends to 1, the value of the waiting time gr also tends to 1 and the probability of selecting a given response interval no longer depends on the distance r. In this case, the probability of obtaining a response interval is equivalent for all the intervals in the size of the contention window, i.e. a probability governed by a uniform random law and equal to 1/W, W denoting the size of the contention window.

Analytical model of the energy consumption of the source device BAT

The source device MTD is in the transmission state T when it transmits the request RR (step F20) or the return message F (step F60), while it is in the reception state R during the contention process. The average energy consumption during the discovery phase is defined by: [Math 17]

where Ts is the duration of a response interval.

The source device MTD transmits its messages to the selected relay device (D2D communication), or directly to the access point BS if no candidate device UE has been found or selected. The average power consumption of the MTD source device during the message transmission phase can be calculated by:
[Math 18]

where Pdisc denotes the probability of discovery, denotes the average energy consumption in D2D communication, and designates the average energy consumption in direct communication with the access point BS.

The average power consumption of D2D communication is defined by:
[Math 19]
where P _m,T designates a constant power of transmission by the source device MTD,
D designates the size of messages sent by the MTD device,
B _eff adjusts the bandwidth efficiency in the NET network,
Ѳ _eff adjusts the efficiency of a signal-to-noise ratio SNR in the NET network,
Bw denotes the transmission bandwidth,
N ₀ denotes the noise power spectral density,
r is the (modified) distance between the source device MTD and the relay device, and
Kd and αd denote the propagation loss factor and exponent respectively for D2D communications.

We study the reduction of the energy consumption of the source device MTD, thanks to the method of the invention, according to an embodiment in which the candidate devices use (step E30) a geometric law truncated by a parameter b.

As already shown, parameter b impacts the discovery probability, the number of response intervals used in the contention process, and the PDF distribution function of the distance between the source devices MTD, and selected relays. All these factors affect the energy consumed by the source device MTD.

Figure 11 illustrates the evolution of the total energy consumption (in the discovery phase and in the message transmission phase) of the source device MTD according to the density of the candidate devices in the discovery zone, according to an example in which the size D of the messages transmitted by the source device MTD is equal to 200 bytes, the radius Rd of the discovery zone is 300 meters, the masking parameter α=8dB, the size W of the contention window is equal to 16 intervals, the distance between the source device MTD and the access point BS is 1000 meters, and the parameter b varies from 0.4 to 1 with a step of 0.2.

FIG. 11 shows that for a low density of candidate devices UE (for λu < 0.8×10 ^-4 devices per m2), the parameter b does not have a significant impact on the power consumption of the source device MTD. However, when the density of the candidate devices UE increases, the parameter b considerably impacts the energy consumption, we notice for example a significant difference between the case where b = 0.4 and the case where b = 1. We also observe that, for this example of implementation, for high densities of candidate devices UE (λu > 0.8x10 ^-4 devices per m2), the value of b = 0.4 is the most appropriate value to minimize the energy consumption of the source device MTD.

FIG. 12 illustrates the evolution of the optimal value of parameter b as a function of the density of candidate devices UE in the discovery zone, according to an example in which the size D of the messages transmitted by the source device MTD is equal to 200 bytes , the radius Rd of the discovery zone is 300 meters, the masking parameter α=8dB, the size W of the contention window is equal to 16 intervals, the distance dm,b between the source device MTD and the point d access BS is equal to 250, 500, 750, or 1000 meters.

We have evaluated the energy consumption for a parameter b which varies from 0.1 to 1 with a step of 0.1, and retained in figure 12 the optimal parameter b which makes it possible to reduce as much as possible the energy consumption of the source device MTD. Figure 12 shows that the lower the density of candidate devices UE, the higher the optimal value of parameter b, and vice versa. For each density, there is a value of the parameter b which minimizes the energy consumption.

Since determining the exact value of the density of UE candidate devices is difficult in real-world scenarios, it is conceivable to use a value of the parameter b that covers the largest range of densities. We consider that for access network scenarios in an urban area and in a dense urban area, the density of candidate devices per square meter is respectively equal to between 0.2x10 ^-4 and 1.2x10 ^-4 . Figure 11 shows that the power consumption of the source device MTD is about the same when the density between 0.2x10 ^-4 and 0.8x10 ^-4 . Moreover, when this density is between 0.8x10 ^-4 and 1.2x10 ^-4 , the optimal value of the parameter b is equal to 0.4. The value of the parameter b can therefore be advantageously chosen in practice in cellular access networks in urban or even dense urban areas.

In one embodiment, the value of the parameter b is between 0.3 and 0.5 for a network with a density of candidate devices between 0.2×10 ^-4 and 1.2×10 ^-4 devices per square meter.

FIG. 13 presents a functional architecture, according to an embodiment of the invention, of a UE candidate device, in accordance with the invention.

The candidate device UE in accordance with the invention can communicate via a channel shared between several devices of the network to respond to the relay request request RR, the relay device being intended to relay messages in the network between a source device MTD and a point access BS of said network.

The UE candidate device comprises:
- A communication module COM configured to receive (E10), from the source device MTD, the relay request request RR;
- a module COM, CONF, for obtaining a size W of a contention window during which the source device is able to receive application messages in response to its request RR;
- a module COM, CONF, for obtaining a threshold value Pth, the communication module COM being configured to send an application message only if a power received from the request RR is greater than or equal to the threshold value;
- a PROC module for obtaining a random waiting time defining a response interval s in said contention window, this PROC module is configured to obtain the waiting time by a mathematical law configured to assign a duration of shorter wait for higher RR request received power;
- a temporal countdown counter TEMP configured to count the waiting time;
- the communication module being configured to listen to the channel; and to send in the channel, if it is free at the end of the waiting period, in response to said request and within the response interval s, a said RC application message indicating that the candidate device UE can constitute a relay device.

In one embodiment, when the threshold value and the size W of the contention window are obtained from the request RR, the module for obtaining these parameters can be the communication module. In another mode, these modules are distinct from each other.

In the embodiments described here, the candidate device UE according to the invention has the architecture of a computer, as illustrated in FIG . It notably comprises a processor 7, a random access memory 8, a read only memory 9, a non-volatile flash memory 10 in a particular embodiment of the invention, as well as communication means 11. Such means are known per se and are not further described here.

The read only memory 9 of the candidate device UE according to the invention constitutes a recording medium in accordance with the invention, readable by the processor 7 and on which is recorded a computer program Prog in accordance with the invention.

The memory 10 of the candidate device UE makes it possible to record variables used for the execution of the steps of the invention, such as the threshold value (Rd and/or Lth and/or a threshold power Pth), the request RR, a power P _RR of reception of the request RR, the propagation attenuation rate Lm,u on the channel between the source device MTD and the candidate device UE, the parameter b, the identifier ID of the device MTD, etc.

The computer program Prog defines functional and software modules, configured to access the shared channel and respond to the RR request. These functional modules are based on and/or control the hardware elements 7-11 of the candidate device UE mentioned above.

Claims (9)

Method for accessing a channel shared between several devices of a cellular access network (NET), called "candidate devices" (UE), to respond to a relay request request (RR), said relay device being intended relaying messages in said network between a communicating device called "source device" (MTD) and an access point (BS) of said network, the source device (MTD) being synchronized with said access point, said method being implemented by a said candidate device, and comprisingsteps of:
- reception (E10) from said source device of said relay request request (RR);
- Obtaining (E25) a size (W) of a contention window during which said source device is able to receive application messages in response to said request;
- obtaining (E30) a random waiting time defining a response interval (s) in said contention window;
- sending in said channel, in response to said request and in said response interval(s), of said application message (RC) indicating that said candidate device can constitute a relay device;
said method being characterized in that it further comprises a step of:
- Obtaining a threshold value, said application message being sent only if a power received from said request (RR) is greater than or equal to said threshold value; and in that
- said waiting time is obtained by a mathematical law configured to produce a shorter waiting time for said higher received power. Access method according to claim 1 wherein said law is a truncated geometric law, the probability that a number s of said response interval in said contention window is chosen by a candidate device being defined by:

where r denotes the distance between the source device and the candidate device obtained from said received power of the request (RR), Rd denotes a distance corresponding to said threshold value, and the parameter b is a constant between 0 and 1 excluded . Access method according to Claim 2, in which the said request comprises the parameter b of the said truncated geometric law. Access method according to one of Claims 1 to 3, in which at least one parameter among said size (W) of the contention window and said threshold value (Rd) is included in said request (RR). Access method according to one of Claims 1 to 4, in which at least one parameter from among said size (W) of the contention window and said threshold value is received from said access point (BS). Access method according to one of claims 1 to 5 further comprising a step of obtaining an identifier of said source device and a step of sending said identifier to said access point with the relayed messages. Computer program (Prog) comprising instructions for the execution of an access method according to one of Claims 1 to 6 when said program is executed by a computer. Computer-readable storage medium on which the computer program (Prog) according to Claim 7 is recorded. Device (UE) of a cellular access network (NET), called "candidate device", able to communicate via a channel shared between several devices of said network to respond to a relay request request (RR), said relay device being intended to relay messages in said network between a communicating device called "source device" (MTD) and an access point (BS) of said network, the source device (MTD) being synchronized with said access point, said candidate device including:
- a communication module (COM) configured to receive, from said source device, said relay request request (RR);
- a module (COM, CONF) for obtaining a size (W) of a contention window during which said source device is able to receive application messages in response to said request;
- a module (PROC) for obtaining a random waiting time defining a response interval(s) in said contention window;
- said communication module being configured to send in said channel, in response to said request and in said response interval (s), a said application message (RC) indicating that said candidate device can constitute a relay device;
said candidate device being characterized in that it further comprises:
- a module (COM, CONF) for obtaining a threshold value (Rd), said communication module (COM) being configured to send said application message only if a power received from said request (RR) is greater or equal to said threshold value; and in that
- said module (PROC) for obtaining a waiting time is configured to obtain said waiting time by a mathematical law configured to produce a shorter waiting time for said higher received power.