FR2934734A1 - Main transmission channel accessing method for communication network, involves accessing main transmission channel during shared time interval in case where applicant node is priority according to predetermined criteria - Google Patents

Abstract

The method involves receiving an access request for accessing a communication network for a shared time interval, where the access request is identical to another access request. The former access request is emitted during collision time interval by an applicant node. A main transmission channel of the communication network is accessed during the shared time interval in a case where the applicant node is priority according to a predetermined criteria from access information contained in the access requests. Independent claims are also included for the following: (1) a computer program product comprising a set of instructions for implementing a method for accessing a main transmission channel of a communication network (2) computer readable storage unit comprising a set of instructions for implementing a method for accessing a main transmission channel of a communication network (3) an applicant node belonging to a communication network, comprising a transmission unit for transmitting an access request.

Description

A method of access by a requesting node to a main transmission channel of a communication network, requesting node, computer program product and corresponding storage means. FIELD OF THE INVENTION The field of the invention is that of communication networks implementing a modulation using cyclic prefixes (or CPs for Cyclic Prefix in English), such as OFDM (for Orthogonal Frequency-Division Multiplexing). in English or orthogonal frequency division multiplexing in French).

Network applications, especially wireless, are now more and more numerous. The need of the users tends towards always more flow of the useful data flow and more quality during the transmission of data. On the other hand, a large number of devices can be brought to communicate with each other and thus constitute a communication network.

To meet the demand for increased data throughput for audio, video, audio and videoconferencing applications, all of which require a low latency communication medium and no loss of quality, it is necessary to increase the bandwidth of the spectrum. radio frequency (or RF for Radio Frequency). Due to physical, normative and legal limitations, it is not possible to increase the bandwidth of the radio frequency spectrum indefinitely. It is then necessary to limit to a certain bandwidth of the radio frequency spectrum around the carrier frequency (a percentage of the carrier frequency for example). For example, a bandwidth of 1% in the authorized band for Wi-Fi (for Wireless Fidelity in English) around the carrier frequency of 2.4 GHz, the bandwidth is 24 MHz. Whereas for a bandwidth of 1% in the authorized band around the carrier frequency of 60 GHz, the bandwidth is 600 MHz. The band of 60 GHz thus allows a greater bandwidth, but also has physical properties different from those of the lower frequencies. One of the physical characteristics of the millimetric band is the low rate of reflection on the obstacles of the type that can be encountered in a domestic environment (walls, furniture, living beings, ....). Practically, this imposes to ensure a good transmission of data without loss, a direct view of transmitting and receiving antennas of communicating objects, in other words an obstacle-free radio wave link except air, and better known under the name of LOS (for Line Of Sight in English). However, as the air attenuates the radio signal, to maintain a good level of quality of the radio link and to have a sufficient radio range without having to emit at unauthorized powers, it is necessary to have directional antennas with a positive gain. . The 60GHz radio transmission systems are particularly well suited for very high speed data transmission in a limited radius, as for example requires a connectivity between different elements of a home cinema (or home theater or home theater in English). Indeed for this use case, the range is limited to ten meters. On the other hand, the bit rates involved are very high, beyond the gigabit per second, due in particular to the nature (audio or video) and the very high resolution of the information transmitted. In addition, still in the case of using a home theater, the transmission system requires perfect synchronism between the transmitter (s) and the receivers, especially in the case of a multi-channel audio system, including up to eight (for example) speakers, and more known as Surround Sound System. Indeed in this specific case, a transmitter (also comprising an audio decoder) perfectly synchronously transmits different audio channels from a single source to a subset of receivers, each comprising a speaker, all of these receivers. to globally restore a multi-spatial sound in a perfectly synchronized manner. Moreover, given the particularly random nature of this type of transmission medium, particularly very sensitive to signal masking caused for example by an individual crossing the transmission field, it is necessary to perform multiple transmissions (via relay) data symbols to ensure good reception beyond a predefined residual error rate. In such an application given here as an example, each loudspeaker of the home theater represents a node of the communication network. After the startup and installation of the system, a number of nodes are chosen to retransmit (ie relay), at time slots allocated according to a defined TDM sequence (for Time Division Multiplexing). English or time division multiplexing in French), the data originally supplied by the source node (the node to which the audio source is connected).

A home theater with 8 speakers, for example a home theater type 7.1, is shown in Figure 1. The listener (30) is surrounded by different nodes or speakers (100, 200, 300, 400, 500 , 600, 700, 800). Each node comprises a transmitting antenna and a receiving antenna (not shown). In Figure 1 are shown the directional beams (351, 451, 551, 651, 751, 851) of each node receiving antenna (100, 200, 300, 400, 500, 600, 700, 800). The transmitting node, during the time interval of the illustrated TDM sequence, is, in this example, the node (100) whose transmission beam (152) is quasi-omnidirectional. The node 100 is here in charge of retransmitting (that is, relaying) the data previously transmitted by the source node (here the node 1000), in order to increase the possibilities for a receiving node to receive at least once the data provided by the source node. The transmission time of a TDM sequence 10 is here shared in several time slots (or time slot in English). In the example of FIG. 1, six time slots IT0, IT1, IT2, IT3, IT4 and C1 compose the TDM sequence. The ITO interval is assigned to the source node 1000 for the first transmission of data from the audio source. The three time slots IT1, IT2, and IT3 allow the nodes to which these intervals have been assigned to transmit on the medium. A node at which a given interval is not affected is in receive mode during that given time interval. To transmit, the nodes use a quasi-omnidirectional antenna, that is to say distributing the radio energy in a large space in order to reach a maximum of nodes. Concerning the nodes in reception, these use directional antennas and directives having the property to have a positive gain (the higher the gain, the more the antenna is directive). This mode of operation requires the selection of the right antenna sector oriented towards the transmitting transmitter during the time slot allocated to it by the determined TDM sequence already mentioned. The different antenna angles to be configured according to the TDM sequence corresponding to the different transmitters can be stored as a table of angles at each node. This angle table can be established during the installation of the complete system by a configuration phase. For example, the nodes perform a spatial scan of each antenna sector and determine the best receive sector for each transmitting node. On a medium such as a 60GHz radio link, shared in TDMA (for Time Division Multiple Access or English Time Division Multiple Access), with predefined time intervals to optimize the use of the bandwidth available, only a few nodes are selected to transmit (data relayed from the source node or data arranged for the first time on the network by the node itself) at determined time intervals (for example time slots IT1, IT2 and IT3 of Figure 1), other time slots being reserved to allow other nodes not having dedicated time slots to request access to the transmission channel and transmit asynchronous data. 2. TECHNOLOGICAL BACKGROUND In order to allow the nodes to request access to the transmission channel in a time interval shared by several nodes (for example the time interval IT4 of FIG. 1), it is customary to reserving a time interval, said collision time interval (or contention time slot in English) such as the time interval C1 of FIG. 1. The collision time interval C1 is then used to transmit transmission requests. access to the medium for a predetermined subsequent IT4 shared time slot. This IT4 shared time slot is associated with this collision time interval C1. Access to the medium during the IT4 shared time slot is non-competitive after processing access requests issued during the interval of time. associated collision time C1 and after determining a requesting node whose access has priority during the shared time slot. Each node, said requesting node, wishing to access the transmission channel during the shared time interval IT4 issues a request during the collision time interval. However, there may be a number of collisions if several nodes issue an access request at the same time, rendering the radio signals received by a receiving node unusable. To decrease the probability of collision, a well-known technique is that the requesting nodes start transmitting in the collision time interval after a randomly determined waiting time for each. Indeed, if there is perfect overlap of transmitted radio frames, the signals received by the receiving nodes are generally unusable. If there is no overlap on the complete frame, but the end of the frame is unusable due to a collision, data will be missing. If it is the beginning of frame that is unusable because of collision, it will also miss data but also essential synchronization information from the beginning of the frame (preamble, header, ...). A logical way to overcome this problem is to re-issue a channel access request as many times as necessary at each new time interval. This has the disadvantage of taking a lot of time and using a lot of bandwidth. This method is used up to a certain limit of a number of re-transmissions, such as in systems of the type CSMA / CD (for Carrier Sense Multiple Access with Collision Detection in English or multiple access with carrier listening and collision detection in French).

This medium access CSMA / CD method is used in standard Ethernet LANs. In this process, each node is free to communicate at any time. Each node sending a message verifies that no other message has been sent at the same time by another node. If this is the case, both nodes wait for a random time before resuming transmission.

In a wireless environment this method is not possible since two transmitting nodes communicating with a receiving node do not necessarily get on mutually due to the range of the radio signal. Thus, the IEEE 802.11 standard proposes a similar protocol called CSMA / CA (for Carrier Sense Multiple Access with Collision Avoidance in English or multiple access with carrier listening and collision avoidance in French). The CSMA / CA protocol uses a collision avoidance mechanism based on a reciprocal acknowledgment principle between the transmitter and the receiver. A node wanting to broadcast listens to the network. If the network is congested, the transmission is delayed. In the opposite case, if the medium is free for a given time called DIFS (for Distributed Inter Frame Space in English or interframe space distributed in French), then the node can emit. The node transmits a message called RTS (for Ready To Send or Request To Send in English), meaning that it is ready to transmit, containing information on the volume of data that it wishes to transmit and the transmission speed. The node to which the data is intended then responds by transmitting a CTS (for Clear To Send in English or free to broadcast in French) meaning that field is free to transmit. Then the sending node starts sending the data. On receipt of all the data transmitted by the sending node, the receiving node sends an acknowledgment ACK (to acknowledge in English). All the nodes of the network then wait for a time which they consider to be that necessary for the transmission of the volume of information to be emitted at the announced speed. This method assumes nodes with omnidirectional antennas. In the case of receiving directional antennas, it is necessary that the antenna of the receiving node is correctly oriented during the RTS request. In the opposite case, the receiving node will not receive the request and therefore the transmitter will have to repeat the request until the antenna of the receiving node which sweeps the space to listen to the requests (or RTS) of the other nodes is correctly oriented to receive the RTS. However, this method is time consuming. It is also a loss of bandwidth. On the other hand, while nodes scan the space to hear if a node is transmitting an RTS, one or more nodes can issue an RTS at the same time. Depending on the position of the directional antenna of the receiving nodes, one or more receiving nodes may receive one or more RTSs, which multiplies the number of possible collision cases and generates a significant unnecessary traffic, resulting in a large number of failures. transmission and making it difficult to exchange data. In addition, in the case of a TDM system, the time interval (s) reserved for this type of traffic are limited in number and duration, which further limits the use of a protocol based on this principle. This method is not enough. A simple method would be to scan all angles corresponding to a direct line of sight (or LOS for Line Of Sight) to each node of the installation in a synchronous manner. When the node is targeted by the directional antennas of the other nodes, it can issue a request for access to the channel, and the process continues so for each node. The disadvantage of this method is that if there is a large number of nodes, scanning all directions to the nodes takes a long time. In this case, the time interval reserved for access requests to the medium may be very long, resulting in a loss of useful bandwidth. Also, this assumes that all the nodes know the position and the angles of sight of all the other nodes, as for example at the start of the system. Many other methods are described in the technical literature and in particular for applications of the wired LAN type, as described in US Pat. No. 6,538,985 (Channel reservation media access control protocol using orthogonal frequency division multiplexing), detailing a way of solving the problem. problem by allocating to each node and their identifier, a frequency in the OFDM spectrum. After the FFT (for Fast Fourier Transform) transformation of the OFDM symbol, the frequency detection serving as an identifier makes it possible to know the nodes requesting access to the medium. This method, which is well adapted to a wired medium, where the attenuation of the signals is weak, is not adaptable to the radio transmission channels, where the attenuations are in this case high. Indeed, a receiving node will pick up a very strong radio frequency RF signal level from the nearby transmitter nodes, and thus adapt the gain in reception to avoid saturation thereof. As a result, the gain will become too small to capture RF signals from more distant nodes. Another invention, as described in patent document W00072626 Method and apparatus for bandwidth allocation, is well suited to the wireless network with quasi-omnidirectional antennas. However, this technique is not suitable for technologies requiring directional antennas requiring orientation at a given time and at a given angle. Indeed, this invention describes a broadband broadband network with nodes called CPE (Customer Premise Equipment in English) and a base station. EPCs may request bandwidth. The requests are controlled by the base station. A first technique described is that of polling (or scanning in French) of one or more CPE periodically or following a request from a CPE. Another technical solution described in this patent document W00072626 is that of Piggybacking, that is to say that the acknowledgment of the previous packet is sent at the same time as the next data packet. This solution also uses the sharing of the medium by a TDD (Time Division Duplex) technique, where time slots are reserved for the CPEs to issue a request. These time intervals are called collision because several CPE can emit at the same time. The patent document W00072626 does not implement techniques for recovering information when collisions occur and the requesting CPE must reiterate its request later, resulting in a loss of time and bandwidth. OBJECTIVES OF THE INVENTION The invention, in at least one embodiment, has the particular objective of overcoming these various disadvantages of the state of the art. More specifically, in at least one embodiment of the invention, one objective is to provide a technique for detecting and selecting in a safe and fast manner a node transmitting a request for access to a transmission channel among a plurality of nodes making a request competitively, even when there are collisions between these requests. At least one embodiment of the invention also aims to provide such a simple technique to implement and inexpensive. 4. DISCLOSURE OF THE INVENTION In a particular embodiment of the invention, there is provided a method of access, by a requesting node belonging to a communication network comprising a set of nodes and clocked by a network cycle broken down into time intervals, to a main transmission channel of said network during a time slot shared by said set of nodes, the requesting node transmitting, during a collision time interval and after a determined duration counted from the beginning of said collision interval , a first access request to said network for the shared time slot, according to the invention, this device is remarkable in that the following steps are performed by said requesting node: - transmitting said first access request, said request access is formulated on an additional channel formed by using cyclic prefixes of data symbols transmitted on the can the main transmission of said network; receiving at least a second access request to said network for the shared time interval, of identical shape to said first access request, and transmitted during said collision time interval by at least a second requesting node; ; accessing said main transmission channel of said network during said shared time slot, in the case where, based on access information contained in said first access request and said second request (s), access, said requesting node has priority according to a predetermined criterion. Thus, in this particular embodiment, the invention is based on a completely new and inventive approach to rapid and efficient selection of a requesting node when there are collisions between at least two access requests to an interval. time share. Indeed, the additional channel obtained (without decoding the symbols themselves, the useful data carried by these symbols on the main channel not being relevant to the present invention), because it uses only the prefix cyclic, allows the receipt of the information frame without synchronization on a preamble (for memory, conventionally, the node must synchronize on a preamble to be able to demodulate data included in the header or in the field of useful data that follows this header ). In the event of a collision, it is not necessary for each requesting node to re-issue a new access request after determining a new pseudo-random duration.

Furthermore, the method of the invention does not require or add little additional hardware to what already exists in a wireless network type. Advantageously, each access request comprises an information frame comprising at least two information fields, said fields containing identical access information. The repetition of the information fields allows a receiving node to increase the chances of recovering useful information when it at least partially receives an access request sent by another node. Indeed, it is sufficient that it receives without collision at least one field of information.

Advantageously, within the information frame, the information fields are separated by separator fields. The separators allow the nodes in receive mode to detect the beginning and the end of the information fields (with the condition that the information field does not contain an identical sequence of bits following the bits constituting the separator).

According to an advantageous characteristic, said access information comprises at least one of the following elements: an identification of the node transmitting said access request; a parameter representative of an occupancy rate of said main channel by the node transmitting said access request; and a parameter representative of the number of accesses prior to said time interval shared by the node transmitting said access request. Thus, by applying an identical process for all the nodes, all the nodes must obtain the same result and thus in the end know the identifier of the node authorized to transmit during the shared time interval. If a masking or fading prevents an access request from being received by all the other requesting nodes access to the medium for the shared time slot, a collision may occur during the shared time slot. However, the risk of such a collision is limited to this situation and most of the collisions occurring during the collision time interval will not prevent the determination of a single node that can access the medium during the split time interval. . According to an advantageous characteristic, each node comprising a reception antenna having a reception coverage defined by an angular spectrum, the duration of each access request is greater than a scanning duration of said angular spectrum by said antenna.

This duration of each access request makes it possible to cover at least one scanning time of all the antenna sectors of the node in reception mode. According to another advantageous characteristic, said duration of each access request is equal to twice said scanning duration.

Because of this constant duration of each access request and the fact that the transmission frames do not start at the same instant, there is a strong probability of existence of at least one window of time where there is no collision between the frames of the different requesting nodes and during which a single node transmits at least part of its access request to the shared time slot.

Advantageously, the step of transmitting said first access request comprises a substep of selecting, for each bit of said first request, between a cyclic prefix value of a data symbol transmitted on the main channel and the inverse of said cyclic prefix value, as a function of the value of said bit of said first access request.

Thus, by simply inverting the cyclic prefix of a symbol, it is possible to transmit additional information of one bit per data symbol transmitted on a main transmission channel and thus create a secondary transmission channel, without bandwidth consumption. additional. In another embodiment, the invention relates to a computer program product downloadable from a communication network and / or recorded on a computer readable medium and / or executable by a processor. This computer program product includes program code instructions for carrying out the aforesaid method (in any one of its various embodiments), when said program is run on a computer.

In another embodiment, the invention relates to a computer readable storage means, possibly totally or partially removable, storing a computer program comprising a set of instructions executable by a computer to implement the aforementioned method (in any of its different embodiments.

The invention also relates to a requesting node belonging to a communication network comprising a set of nodes and clocked by a network cycle broken down into time slots, said requesting node requesting access to a main transmission channel of said network during a time interval. shared by said set of nodes, the requesting node transmitting, during a collision time interval and after a determined duration counted from the beginning of said collision interval, a first access request to said network for the shared time slot. According to the invention, this requesting node is remarkable in that it comprises: means for transmitting said first access request, said access request being formulated on an additional channel formed by using cyclic prefixes of data transmitted on the main transmission channel of said network; means for receiving at least a second access request to said network for the shared time slot, of a form identical to said first access request, and transmitted during said collision time interval by at least a second requesting node; means for accessing said main transmission channel of said network during said shared time slot, in the case where, based on access information contained in said first access request and said second request (s) ( s) access, said requesting node has priority according to a predetermined criterion. Advantageously, each access request comprises an information frame comprising at least two information fields, said fields containing identical access information. Advantageously, within the information frame, the information fields are separated by separator fields. According to an interesting characteristic, said access information comprises at least one of the following elements: an identification of the node transmitting said access request; a parameter representative of an occupancy rate of said main channel by the node transmitting said access request; and a parameter representative of the number of accesses prior to said time interval shared by the node transmitting said access request. In addition, each node comprising a receiving antenna having a reception coverage defined by an angular spectrum, the duration of each access request is greater than a scanning time of said angular spectrum by said antenna. Also, said duration of each access request is equal to twice said scanning time. According to another advantageous characteristic, the transmission means comprise selection means, for each bit of said first request, between a cyclic prefix value of a data symbol transmitted on the main channel and the inverse of said cyclic prefix value, as a function of the value of said bit of said first access request. 5. LIST OF FIGURES Other features and advantages of the invention will appear on reading the following description, given by way of indicative and nonlimiting example, and the appended drawings, in which: FIG. 1 schematically illustrates, in the space domain and at a time interval IT1 of a super frame 10 of a TDM cycle, the configuration of a system of eight nodes according to a particular embodiment of the invention; FIG. 2 schematically illustrates, in the spatial domain and at a time interval IT2 of the superframe 10, the configuration of the eight-node system of FIG. 1; FIG. 3 schematically illustrates, in the spatial domain and at a time interval IT3 of the superframe 10, the configuration of the eight-node system of FIG. 1; FIG. 4 schematically illustrates, in the spatial domain and at a collision time interval C1 of the superframe 10, the configuration of the eight-node system of FIG. 1; FIG. 5 schematically illustrates, during the collision time interval, the situation of the receiver antennas of the receiving nodes of an access request, sent by a requesting node 700, to a shared time interval IT4 of the superframe. 10; FIG. 6 diagrammatically illustrates, during the collision time interval C1, the situation of the receiving antennas of the receiving nodes of two simultaneous access requests transmitted by two requesting nodes 600 and 700; FIG. 7 schematically illustrates, during the collision time interval C1, three RF frames of information each corresponding to a request for access to the shared time slot IT4 and issued by three requesting nodes; FIG. 8 schematically illustrates the creation of an additional channel carrying a request for access to the shared time slot IT4, according to a particular embodiment of the invention; FIG. 9 schematically illustrates a particular arrangement of an information frame of an access request transmitted by a requesting node, according to a particular embodiment of the invention; FIG. 10 schematically illustrates, in the spatial domain and at the shared time interval IT4 of the superframe 10, the configuration of the eight-node system of FIG. 1 when access to the medium has been assigned to a node 700 , which has previously requested it during the collision time interval C1; - Figure 11 schematically illustrates a block diagram of a device according to a particular embodiment of the present invention; FIG. 12 illustrates an algorithm for managing access to a shared time slot IT4 transmitted by a requesting node, according to a particular embodiment of the invention. 6. DETAILED DESCRIPTION The following description illustrates a particular embodiment of the invention. Several starting hypotheses are to be considered in connection with FIG. 1. - the set of nodes of the communication network are synchronized on a predetermined TDM sequence at the start of the type 7.1 home-cinema system in this particular embodiment; the configuration of FIG. 1 comprises eight nodes and a super-frame of fixed duration comprising six time slots: the first ITO time slot is allocated to the transmission of data by the source node 1000; O the following three time intervals IT1, IT2, IT3 are allocated to three other nodes among the eight nodes; the fifth time slot corresponds to a collision time interval Cl (or contention time slot in English), where the nodes (in particular those having no allocated transmission time slots 30 among the ITO to IT3 intervals ) can issue a request to access an IT4 shared time slot; the sixth time interval corresponds to a shared time interval IT4, and is allocated to the node whose access request transmitted during the collision time interval C1 has previously been satisfied; The home theater system of type 7.1 of FIG. 1 is operational, that is to say that it has been configured: each node knows the value of the antenna angles to be used to receive the signal radio nodes that transmit at the first four time intervals ITO to IT3, but do not necessarily know the antenna angles necessary to receive the radio signal from other nodes requesting access to the communication network (medium); each node has two types of antenna: a first omnidirectional transmission antenna, and a second antenna-type antenna (or beam-forming antenna) allowing different configurations of the antenna lobe. reception, that is to say a narrow lobe with a positive gain. FIG. 1 represents, in the spatial domain and at a time interval IT1 of a superframe 10, the configuration of a home-theater system comprising 8 nodes (or loudspeakers) which may be, for example, a type of audio system. 7.1. The position of the time interval IT1 is indicated by a representation (10) of the TDM cycle on a super-frame, the gray part corresponding to the current time interval. The time interval ITO during which the node 1000 has sent data on the communication network has therefore elapsed for the current TDM cycle. The node 100 transmits for the duration of the time interval IT1 thanks to its omnidirectional antenna. This transmission, represented by the radio transmission lobe (152) of the transmitting antenna of the node 100 is performed to all other receiving nodes (200, 300, 400, 500, 600, 700, 800). The receiving lobes (251, 351, 451, 551, 651, 751, 851) of the directional antennas (not shown) of the nodes (200, 300, 400, 500, 600, 700, 800) then orient towards the node 100 transmitter. Each reception lobe represents the orientation of the directional antenna of the node in reception mode. The circle 35 represents the placement of the loudspeakers around a listener 30 placed substantially in the center of the home cinema system, according to the ITU-R BS.775-2 recommendations, given here by way of example, the invention being operable to any other configuration.

In this case of application, the nodes (100, 200, 300, 400, 500, 600, 700, 800) are loudspeakers with speakers which respectively have the function of central speaker (central speaker) , left side speaker, right side speaker, front right speaker, front left speaker (front left speaker in English), rear left speaker, rear right speaker, and subwoofer. The arrow 31 represents the abscissa axis of the Cartesian coordinate system (x, y) for indicating the orientation of the directional antennas of each acoustic enclosure or node. FIG. 2 shows the situation of the reception lobes at a time interval IT2 of the superframe 10, which is assigned to the data transmission by the node 200. During this time interval, only the node 200 transmits, with its quasi-omnidirectional transmission antenna (represented by its emission lobe 252). The other nodes, in receive mode, in this case orient their receive directional antennas to the transmitter node 200. FIG. 3 shows the situation of the receive lobes at a time interval IT3 of the superframe 10, which is assigned to the data transmission by the node 300. During this time interval, only the node 300 transmits, with its quasi-omnidirectional transmission antenna (represented by its emission lobe 352). The other nodes, in receive mode, in this case orient their receive directional antennas to the transmitter node 300. FIG. 4 represents the reception lobes situation at a collision time interval C1 of the superframe 10, which is dedicated to the transmission of access requests to a shared time interval IT4 of the superframe 10. During this collision time interval C1, all the nodes wanting to access the medium can send an access request to an IT4 shared time slot (more fully described later). Thus, the nodes to which it has not been assigned a time interval among the time intervals ITO to IT3, or who wishes momentarily to obtain additional bandwidth, can obtain access to the medium during the time interval IT4 by transmitting, during the collision time interval C1, an access request. Each requesting node (also called the requesting node), outside the time it is required to issue its request, scans the space per antenna sector as represented by the dotted antenna lobes (151, 251, 351 , 451, 551, 651, 751, 851), in search of a signal corresponding to a concurrent request sent by one of the other nodes of the network. As illustrated in FIG. 5, if a requesting node 700 issues an access request (represented by its transmission lobe 751), the reception mode nodes receive the radio signal thereof when their antenna is passed through the port. direction of the requesting node (node 700 in this case). They then stop scanning the antenna sectors and freeze the antenna in the direction of the node 700. After having captured the radio signal and decoded the useful data (steps described in more detail in the following description), each node in Receive mode records this data. Once these data are stored, the nodes restart a spatial scan of the different antenna sectors in order to look for other nodes issuing an access request to the IT4 shared time slot in order to access the medium. At the end of the collision time interval C1, the reception mode nodes determine (as described in more detail in connection with FIG. 12), from the stored information, the node authorized to transmit during the interval. IT4 time share. FIG. 6 illustrates several nodes transmitting simultaneously during the collision time interval C1. In the configuration of FIG. 6, the nodes 600 and 700 simultaneously transmit an access request to the medium (represented by means of FIG. their respective emission lobes 651 and 751). In this case, the RF signals emitted by the nodes 600 and 700 are likely to disrupt, cancel or mutually abate, preventing or not guaranteeing the integrity of the RF signals. The invention solves this problem as illustrated in the following figures. Figure 7 schematically illustrates a super frame 10 comprising four time slots 11 assigned to four nodes. These time slots serve for example for the transmission of synchronous data. The superframe 10 also includes a collision time slot 20 for access requests to a shared time slot 12. In the example of FIG. 7, three nodes issue competitively during the collision time interval. 20 access requests to said shared time slot 12. These access requests are represented by three frames of RF information 22, 23 and 24. According to the invention, so that these RF information frames are not perfectly aligned in time, an offset in time at startup of these frames is implemented. This makes it possible in particular to avoid the perfect recovery of an information frame by another information frame. Indeed, each frame must have, in order to be correctly received by other nodes, a time interval during which it is the only one emitted on the medium. Otherwise, it can not be perceived by the other nodes (receiving nodes) and the transmitting node will not be aware of it (in such a configuration, a transmitting node can not receive a frame at the same time). In the example of FIG. 7, the invention proposes to temporally shift the three RF frames 22, 23 and 24, respectively, of a duration a, b, and c determined pseudo randomly from the beginning of said collision time interval. However, if two nodes transmit exactly at the same time despite a determined pseudo random start time for each of the two nodes and their transmission time intervals are perfectly aligned, then it is impossible to decode the RF information frames by In the receive nodes, it is possible for several nodes to consider themselves, at the end of the collision time interval 20, as having obtained the right of access to the medium for the shared time slot 12. In this case, there will be collision during the shared time slot 12; it will then be necessary to wait for the next superframe and restart the procedure, with normally different pseudo-random start times. To overcome this problem, it is chosen a constant duration of the sending frames of the requesting nodes as described below. It should be noted that this situation also occurs when a significant masking or fading occurs during the collision time interval 20, thus preventing the broadcast of all the access requests to all the nodes (it At least the requesting nodes must correctly receive the access requests issued by the other nodes to avoid subsequent collisions during the shared time interval 12). In such a case, retransmission of the access request during a next collision time interval is necessary.

The graduations 21 represent a duration d corresponding to the maximum scanning time of all the antenna sectors of the communication network by a receiving node. The scanning time of the entire angular spectrum of coverage of the receiving antenna is determined by the number of sectors of angles to be scanned, the response time of the RSSI (for Received Signal Strength Indication), the time of reception. analysis of this RSSI and subsequent decision, to which must be added the consideration that the analysis must be done on at least three OFDM symbols, to confirm that the detected radio signal is from a node belonging to the system. The duration of the sending frames of the requesting nodes is constant. In particular, the length of the frame transmitted during the collision time interval 12 is a function of the scan time of the angular spectrum. Indeed, a node that has just switched to reception mode, begins by scanning with its directional antenna the space by analyzing for each antenna sector the strength of the received radio signal (RSSI). Once it has scanned all antenna sectors without finding a radio signal, it starts scanning again until it detects a radio signal. If this node started before a requesting node emits a signal at time Tk + 1 (i.e., when the receiving node has arrived at antenna sector k + 1), and it has already scanned the antenna sector k in which transmits the requesting node, it must be able to recover the signal at the next scan that is to say when it will arrive at the antenna sector k. That is why the sending duration of the requesting node must cover more than one scan time of all antenna sectors of the node in receive mode. It is calculated from a number n of sweeps of the whole angular spectrum. In a preferred embodiment, this number n must be strictly greater than 1. For example, n takes the value 2, ie a transmission duration equal to 2 * d.

Due to this constant duration of the sending frames of the requesting nodes, and the fact that the transmission frames do not start at the same instant, there is a strong probability of existence of at least one time window, represented here. by fl, f2 or f3, where there is no collision between the RF information frames of the different requesting nodes and during which a single node transmits at least a portion of its access request to the time slot sharing. Each constant duration is at least equal to or greater than the time interval 21 of duration d corresponding to the complete scanning of the space by the antennas. The other nodes which are at this time in reception mode then pick up the signals of a transmitting node for a constant duration. In addition, the starting times a, b or c of a frame of a requesting node are a submultiple of the scanning time d of the whole of the angular spectrum. The duration of transmission of the requesting node is fixed, as explained above at 2 * d, for example. When there are overlaps of several request node frames, it is necessary that the non-overlapping area, for example f2, be non-zero so that a receiving node can have a chance to pick up a signal, even if a full scan of Antenna areas is not possible. These random delays can take for example the values k * d / 2 or k * d / 3 where k is an integer of 0 up to a maximum value corresponding to the total duration of the collision time interval, minus the duration of a transmission frame (for example as chosen in the example above at 2 * d).

This makes it possible to have a window f2 of minimum duration allowing the sweeping of judiciously selected sectors (for example with a history of sectors that have already been active), and thus to capture with a higher probability a transmitter. In another embodiment, these delays are completely random with however limited amplitude values.

Obviously, the values presented are as a non-limiting example.

However, there is a high probability that a time window (for example f2) does not allow the receiving nodes to capture the beginning and / or end of an information frame (for example the frame 22 of FIG. ), and thus, to synchronize on the preamble to decode the information of each access request contained in the header of the corresponding information frame carried on a transmission channel, said first main channel. To overcome this problem, the present invention suggests a particular arrangement of data using the notion of cyclic prefixes to create an additional channel on which the access request is transmitted. The patent document FR 2845842 details this technique of generating an additional communication channel by using the cyclic prefixes inserted in the communication systems using the OFDM modulation. Thus, this document relates to the field of wireless communication systems and, more specifically, communication systems implementing a TDMA-FDMA type structure (for Time Division Multiple Access and Frequency Division Multiple Access in English). These communication systems can be either highly centralized, ie have a base station that manages the entire network and allocates access to the networks to the various members, or more distributed, that is to say that network access decisions are made by one or more terminals that decide for themselves or for others access rights to the network.

These communication systems have the advantage of having a main communication channel (broadband) for carrying useful information related to the user (called payload data in English) and a secondary communication channel for management information. network, also called guide channel (English beacon). The secondary communication channel is normally managed by the base station, which is in principle connected to a wired network of telephony or data type. A terminal is capable of transmitting a certain amount of information to the base station describing its own location with respect to the network. The generation of the secondary communication channel is obtained by using the cyclic prefixes inserted in the communication systems implementing the OFDM modulation. By modifying the cyclic prefix adequately, it is possible to carry additional information per OFDM symbol without changing the main characteristics of the system usually used. Indeed, the cyclic prefix is an extension of the OFDM symbol made after the modulation operation commonly performed by a Fast Fourier Transform (FFT). Classically, this extension is performed by copying, in the header of each OFDM symbol, a certain number of samples from the end of the symbol. This operation has the advantage of maintaining the orthogonality of the signal during the useful symbol and thus of solving inter-symbol interference problems related to the time dispersion of the signals received at the input of the receiver. The size of the cyclic prefix is therefore to be determined according to the radio channel and the expected maximum time dispersion. Another advantage of the cyclic prefix is to assist the time synchronization process of the receiver and in particular the recovery of the symbol clock. Indeed, the redundancy introduced makes it possible to use an auto-correlator to find a signal whose frequency will correspond to the symbol clock used in the transmitting node.

The advantages of this known technique are, on the one hand, the creation of an additional communication channel without any loss of flow, since no main communication channel (allowing the transmission of useful data) is effectively sacrificed. On the other hand, the receiving terminal network management information can extract the data from the secondary channel very simply, without having to demodulate the data received and thus saving energy. Note that in the case where the receiving node already uses an autocorrelator to perform symbol synchronization, an additional comparator is sufficient. In the opposite case, it is necessary to add also the auto-correlator, which nevertheless remains less expensive compared to an operation of demodulation by FFT.

In addition to the means implemented in a conventional OFDM system, the technique described in this document requires the following means implemented by the transmitting node: means for modifying the polarity of the samples of each cyclic prefix according to the management data the network to be transmitted; And by the receiver: means for determining (1102, 1103, 1105, 1106) an auto-correlation signal on the received signals; - Comparison means (1108, 1109) of the value of the signal from the autocorrelator with a reference value for deriving a value of the network management data that had been coded. The operation of the system for generating an additional channel can be summarized as follows. On transmission, the cyclic prefix is determined conventionally (copying the last 64 complex points of the OFDM symbol at the beginning of the symbol, for example), but its polarity is modified according to the management data of the channel that is desired. transmit for the symbol concerned. For example, it can be decided that information "1" will not change the cyclic prefix and that information "0" will reverse its value. The data being complex, the signs of the real component and the imaginary component of the prefix samples can therefore be modified. On reception, a conventional autocorrelation function is applied and the management data can be extracted as follows: if the amplitude of the real part of the autocorrelation signal applied to the cyclic prefix is positive, we deduce that the first bit is "0", otherwise it is "1". The management data from the secondary channel thus extracted are then transmitted to the network management unit which operates in a conventional manner. The operation of all other parts of the OFDM system is also standard. Thus, the receiving node capturing an RF information frame can notably use auto-correlators for detecting these cyclic prefixes. By reusing this principle, and transporting 1 bit by cyclic prefix, that is to say by symbol, the receivers can decode information contained in the cyclic prefixes without having to capture the beginning or the end of the information frame. RF. This principle is more fully illustrated in connection with FIG. 8. A conventional frame 8002 consists of a preamble, a header and a field for payload 8003 (or payload). It is in this payload field 8003 that are the different symbols modulated by the data, each symbol corresponding to a part called 8004 cyclic prefix used in the OFDM modulation to frame the type of transformation FFT (Fast Fourier Transform in English).

Figure 8 details an arrangement 8005 of all these cyclic prefixes 8004 to carry a bit to each symbol and thus benefit from an additional 8006 low rate channel. The invention proposes to reuse this type of additional channel by specifically ordering the bits to form an additional channel 8007 on which a requesting node transmits, during the collision time interval 20, its access request to the shared time slot 12. The access request comprises in particular an information frame 9000 in which is indicated at least one information field 9001 separated by a separator 9002. FIG. 9 describes this particular arrangement for a frame of information 9000 issued by a node requesting access to the medium for the shared time slot 12. It is composed of an information field 9001 repeated here 5 times (for example) in the frame and separated by 9002 separators. Each information field 9001 contains access information (9003, 9004) for determining the priority level of each node requesting access to the medium. The first information contained in the field 9003 informs on 8 bits the occupancy rate of the medium by the node. The priority of a node for access to the medium, among all the other nodes requesting access to the medium, is determined as a function, directly or indirectly, of the information as expressed in the field 9003 (by for example, the access is assigned to the node for which the field 9003 of the access request is the smallest). Other criteria for determining the node obtaining access to the medium during the shared time slot 12, from queries issued during the collision time interval, can be used, such as for example the number of previous accesses. at the shared time slot 12 by the node issuing the request. In addition, these parameters can be weighted.

The second information contained in the 9004 field informs on 6 bits the identifier of the sending node of the access request to the medium. Thus, each receiving node can from this information determine which node is requesting. This field 9004 further allows, in case of equality on the number of accesses (field 9003), to give priority to a node having the value of the field 9004 the weakest.

A third piece of information contained in a field 9005 informs on 2 bits a Cyclic Redundancy Check (CRC) value to make it possible to check the validity of the received information. The separators serve for the nodes in the reception mode to detect the beginning and the end of the information fields 9001, with the condition that the information field 9001 does not contain a sequence of bits identical to the separator. ie the value 0110 in the example of FIG. 9. The repetition of the information fields 9001 (5 times in the example of FIG. 9) allows a receiving node receiving an RF signal coming from a requesting node, despite a narrow time window, to retrieve useful information. The content of the frame transmitted during the collision time interval C1 of FIG. 1 by a requesting node is a non-zero signal and has the characteristics enabling the detection electronics to be better calibrated. But this content does not necessarily contain useful data. In a particular embodiment, these data are a fixed OFDM pattern (or pattern) of short symbol type (or short symbol in English) with a duration of about 100ns, allowing the self-correlator to receive optimum operation. FIG. 10 schematically illustrates the transmission, during the shared time interval IT4 of the superframe 10, of a node 700 authorized to transmit after execution of an algorithm for selecting a requesting node from among a set of other nodes requesting a request for access to the medium. During this time interval, only the node 700 transmits, with its quasi-omnidirectional transmission antenna (represented by its transmission lobe 751). The other nodes, in receive mode, in this case orient their receive directional antennas to the transmitter node 700.

FIG. 11 diagrammatically illustrates a block diagram of one of the possible hardware implementations of a node embodying the invention. Conventionally, for a wireless system, each node of the invention comprises a radio part with a receiver module 1100 (denoted Rx in FIG. 11) and a transmitter 1119 (denoted Tx in FIG. 11), interfaced to a trunk circuit. base 1120 (or baseband in English) which itself is interfaced to a MAC layer 1115 (for Medium Access Control in English). In the baseband circuit 1120 are the analog-digital converter 1101 (or ADC for Analog-to-Digital Converter in English) on the receiving side and the digital-analog converter 1118 (or DAC for Digitalto-Analog Converter in English) on the emission side .

Also conventionally, there is integrated a modem 1117 of OFDM type interfaced by a bus 1116 to the MAC layer 1115. After processing by the circuit 1101, the data are sent directly by means of a link 1104 to the modem 1117, and also to blocks 1102, 1103, 1105, 1106, 1107, 1108, 1109 representing the self-correlator and whose operation is more fully described in patent FR 2845842.

At the output 1110 of this autocorrelator, the control bits are then sent to a microcontroller 1112 conventionally equipped with memories 1111 of RAM (for Random Access Memory in English) and ROM (for Read Only Memory) type and which loads the different steps of the algorithm for selecting a requesting node from among all the requesting nodes, to grant it access to the medium during the shared time slot, and more fully described in relation with FIG. 12. Figure 12 illustrates the algorithm for selecting a requesting node from among all the requesting nodes to grant access to the medium during the shared time slot. The algorithm is implemented in any node of the network (receiving node or transmitter, requesting or not), said current node. Conventionally, a time base integrated at each node makes it possible to know the timing of each TDM cycle in order to synchronize at the beginning of a collision time interval C1 (called current collision time interval) of the superframe 10 FIG. 1. The algorithm starts at a step 1200. At this precise start time of the current collision time interval, different indices are set to zero: • t: current time indication, since the beginning of the current collision time interval until the end of the current collision time interval; • x: an indication corresponding to the sectors of antenna angles and taking the values of 0 to a maximum value x final; • s: emission index of the access request taking the value 1 if the access request is made or not to be done and taking the value 0 if the access request is to be made. • a: delay before sending, by the current node when it is a requesting node, a request for access to the medium for a subsequent IT4 shared time slot. In a step 1201, the need for the current node to access the medium during a subsequent IT4 shared time slot is determined. In a step 1202a, if an access request is to be sent during the collision time interval C1, then the subscript s has the value 0. Then, in a step 1212, the value of the delay a is determined in a pseudo manner. -random. If no access request is to be sent during the collision time interval C1, then in a step 1202b, the subscript s takes the value 1 and the delay a takes the value 0.

Steps 1202b and 1212 are both followed by a step 1203 of scanning the different antenna sectors in order to look for a node transmitting a request for access to the medium for a subsequent IT4 shared time slot. Then in a step 1204 is checked whether an RF signal is detected by the current node executing the present algorithm. If a splitter is detected, in a step 1213, the receiving system of the node decodes and stores the data received from that requesting node before performing a step 1205 of checking whether the current node (if it desires access to the medium also ) is allowed to transmit, that is, if the time indication value t is greater than the value of the delay a. If no separator is detected in step 1213, this means that the detected signal is not a signal transmitted according to the communication protocol implemented in the network, and step 1205 is directly executed. In step 1205, if the value of the index t is smaller than the calculated value a, then another antenna sector is activated (step 1203) and is listened again (steps 1204, 1213). A step 1217 makes it possible to increment in this case by one unit the value of the index x. If the time indication value t is equal to or greater than the delay value a, a step 1206 makes it possible to check whether the current node has sent a request for access to the medium for a subsequent IT4 shared time slot. If in step 1206 the node has not issued a request (s = 0), then the current node is allowed to issue a request to access the medium in a step 1214 before executing a step 1207.

If in step 1206 the node has already issued a request or if a request is not to be made (s = 1), step 1207 is directly executed. In step 1207, the current node determines whether time indication value t is greater than or equal to the duration of the collision time interval. If this is not the case, the current node continues the scanning process of the antenna sectors by incrementing each time the index x of the antenna sector through the consecutive steps 1217b, 1203b, 1204b and 1213b respectively in steps 1217, 1203, 1204 and 1213. Then, in a step 1215 is checked whether the value of the index x, after a complete scan of all the antenna sectors, is equal to an xfinal value representing the maximum number of antenna sectors.

If this is the case, the current node resumes a complete scan by initializing the value x to 0. If this is not the case, all the steps from step 1203 are executed again.

If the time indication value t is greater than the duration of the collision time interval (step 1207), the current node compares, in a step 1208, the information (step 1213) corresponding to the other requesting nodes with its own values (if the current node is also one of the applicants). In case of equality of the medium occupancy rate or the number of accesses (field 9003 of FIG. 9), the logic of the current node compares the identifiers. The decision is made on the highest or the lowest value of the occupancy rate as explained above. This process is the same for all the nodes, thus guaranteeing that all the nodes obtain the same result and thus in the end know the identifier 9004 of the node authorized to transmit during the shared time interval IT4 of FIG.

After executing step 1208, in a step 1209, the current node compares its own identifier with the value of the authorized identifier to be transmitted during the shared time interval IT4. If these two values are not identical, the current node waits for the next superframe and a new collision time interval as indicated in a step 1211. If these two values are identical, in a step 1210, the current node is allowed to transmit data during the IT4 shared time slot. When a new collision time interval arrives (step 1211), the algorithm is re-implemented (step 1200).

It will be noted that the particular embodiment of the invention has a collision time slot and an associated shared time slot belonging to the same super frame. Note that the invention is not limited to this particular embodiment and that the shared time slot associated with a collision time interval and this collision time interval can be in distinct super frames, and the number of cycles separating these super frames being predefined, for example by construction of the nodes of the network. It will be noted that the invention is not limited to a purely material implantation but that it can also be implemented in the form of a sequence of instructions of a computer program or any form mixing a material part and a part software. In the case where the invention is partially or totally implemented in software form, the corresponding instruction sequence can be stored in a removable storage means (such as for example a floppy disk, a CD-ROM or a DVD-ROM) or no, this storage means being partially or completely readable by a computer or a microprocessor.

Claims (16)

REVENDICATIONS1. A method of accessing, by a requesting node belonging to a communication network comprising a set of nodes and clocked by a network cycle decomposed into time slots, to a main transmission channel of said network during a shared time slot (IT4) by said set of nodes, the requesting node transmitting, during a collision time interval (C1) and after a determined duration (a, b, c) counted from the beginning of said collision interval, a first access request to said network for the shared time interval (IT4), the method being implemented by the requesting node, characterized in that the method comprises the following steps: - transmitting (1214) said first access request, said request for access being formulated on an additional channel (8007) formed by using cyclic prefixes of data symbols transmitted on the main transmission channel of said network; receiving (1213, 1213b) at least one second access request to said network for the shared time interval (IT4), of identical shape to said first access request, and transmitted during said time interval of collision by at least a second requesting node; accessing said main transmission channel of said network during said shared time interval (IT4), in the case where, based on access information contained in said first access request and said second request (s) ( s) access, said requesting node has priority according to a predetermined criterion. 2. Method according to claim 1, characterized in that each access request comprises an information frame (9000) comprising at least two information fields (9001), said fields containing access information (9003, 9004) identical. 3. Method according to claim 2, characterized in that, within the information frame (9000), the information fields are separated by separator fields (9002). 4. Method according to any one of claims 1 to 3, characterized in that said access information comprises at least one of the following elements: - an identification (9004) of the node transmitting said access request; a parameter (9003) representative of a occupancy rate of said main channel 35 by the node transmitting said access request; and 20- a parameter (9003) representative of the number of accesses prior to said time interval shared by the node transmitting said access request. 5. Method according to any one of claims 1 to 4, characterized in that, each node comprising a receiving antenna having a reception coverage defined by an angular spectrum, the duration of each access request is greater than a duration. scanning (d) of said angular spectrum by said antenna. 6. Method according to claim 5, characterized in that said duration of each access request is equal to twice said scanning time (d). 7. Method according to any one of claims 1 to 6, characterized in that the step of transmitting said first access request comprises a substep of selecting, for each bit of said first request, between a value cyclic prefix of a data symbol transmitted on the main channel and the inverse of said cyclic prefix value, as a function of the value of said bit of said first access request. 8. Computer program product downloadable from a communication network and / or recorded on a computer readable medium and / or executable by a processor, characterized in that it comprises program code instructions for the implementation of the method according to at least one of claims 1 to 7, when said program is run on a computer. 9. Computer-readable storage means, possibly totally or partially removable, storing a computer program comprising a set of instructions executable by a computer to implement the method according to at least one of claims 1 to 7. An applicant node belonging to a communication network comprising a set of nodes and clocked by a network cycle decomposed into time slots, said requesting node requesting access to a main transmission channel of said network during a shared time slot (IT4) by said set of nodes, the requesting node transmitting, during a collision time interval (C1) and after a determined duration (a, b, c) counted from the beginning of said collision interval, a first access request to said network for the shared time slot (IT4), characterized in that said requesting node comprises: means for transmitting said first access request, said access request being formulated on an additional channel (8007) formed by use cyclic prefixes of data symbols transmitted on the main transmission channel of said network; means for receiving at least one second request; said network access port for the shared time slot (IT4), of identical form to said first access request, and transmitted during said collision time interval by at least a second requesting node; means for accessing said main transmission channel of said network during said shared time interval (IT4), in the case where, based on access information contained in said first access request and said second one ( s) access request (s), said requesting node has priority according to a predetermined criterion. The requesting node according to claim 10, characterized in that each access request comprises an information frame (9000) comprising at least two information fields (9001), said fields containing access information (9003, 9004) identical. 12. Requesting node according to claim 11, characterized in that, within the information frame (9000), the information fields are separated by separator fields (9002). 13. Requesting node according to any one of claims 10 to 12, characterized in that said access information comprises at least one of the following elements: an identification (9004) of the node transmitting said access request; a parameter (9003) representative of a occupancy rate of said main channel by the node transmitting said access request; and a parameter (9003) representative of the number of accesses prior to said time interval shared by the node transmitting said access request. 14. Requesting node according to any one of claims 10 to 13, characterized in that, each node comprising a receiving antenna having a reception coverage defined by an angular spectrum, the duration of each access request is greater than one. scanning duration (d) of said angular spectrum by said antenna. 15. Requesting node according to claim 14, characterized in that said duration of each access request is equal to twice said scanning time (d). 16. Requesting node according to any one of claims 10 to 15, characterized in that the transmission means comprise selection means, for each bit of said first request, between a cyclic prefix value of a transmitted data symbol. on the main channel and the inverse of said cyclic prefix value, as a function of the value of said bit of said first access request.