Patent Application
20080316052
The present invention relates to a controller unit, in particular to a central data processing unit, for example to a relay control box, as well as to a method for controlling communication between and among mobile nodes, in particular between and among vehicles, each node being designed for receiving and transmitting messages, in particular
The present invention further relates to a corresponding communication device for communication between and among mobile nodes, in particular between and among vehicles, as well as to a communication system for wireless L[ocal]A[rea]N[etwork]s for communication between and among mobile nodes, in particular between and among vehicles.
The present invention further relates to a communication protocol for controlling communication between and among mobile nodes, in particular between and among vehicles, each node being designed for receiving and transmitting messages, in particular
The choice of data rate and of transmission power for wireless local area networks (so-called wireless LANs or WLANs) is still an open problem. In fact so far these types of network are mainly used to connect multiple stations to a central access point. In this situation the best choice is to transmit with the highest data rate possible at the highest power available.
The reason is that the access point represents a “bottleneck”, and the choice of the highest data rate minimizes the time in which the access point is busy. The interference does not represent a main problem in this case because during the time the access point is busy, the nodes are not allowed to transmit to other nodes (in the standard operating mode).
A road safety wireless LAN has to function also without an access point. This implies that the mobile nodes mainly exchange messages with each other, and only occasionally connect to a fixed access point.
Under these conditions the choice of the highest data rate at the maximum power does not represent anymore the best solution because this implies a high level of interference preventing other nodes from exchanging messages.
In the prior art literature some proposals of data rate and power control selection mechanisms can be found. In prior art article “Efficient Power Control via Pricing in Wireless Data Networks” by Cem U. Saraydar, Narayan B. Mandayam, and David J. Goodman, IEEE Transactions on Communications, volume 50, issue 2, February 2002, pages 291 to 303, the concept of pricing function is introduced, which assigns a cost depending on the power used for every transmission. An algorithm is finally proposed minimizing this cost function for every node and giving a more equal distribution of the bandwidth.
In prior art article “Power controlled multiple access (PCMA) in wireless communication networks” by Nicholas Bambos and Sunil Kandukuri, INFOCOM 2000, 19th Annual Joint Conference of the IEEE Computer and Communications Societies, Proceedings, IEEE, volume 2, Mar. 26 to 30, 2000, pages 386 to 395, an algorithm regulating transmission power for every node depending on the number of messages in the queue waiting to be transmitted and on the power of the interference detected is proposed.
The results show that this algorithm achieves higher throughput compared to a standard constant signal-to-interference (so-called constant SIR) algorithm where the power is increased depending on the bit error rate calculated at the receiver.
Moreover, in prior art article “Multimodal Dynamic Multiple Access (MDMA) in Wireless Packet Networks” by Sunil Kandukuri and Nicholas Bambos, INFOCOM 2001, 20th Annual Joint Conference of the IEEE Computer and Communications. Societies, Proceedings; IEEE, volume 1, Apr. 22 to 26, 2001, pages 199 to 208, an extension of the algorithm is proposed also including the choice of data rate, beside the selection of the transmission power. Results show a further increase in the network throughput.
Regarding “reinforcement learning” reference can be made for example to prior art article “Reinforcement Learning: A Survey” by L. P. Kaeblling, M. L. Littman, and A. W. Moore, Journal of Artificial Intelligence Research 4, 1.996, pages 237 to 285. The main problem of these algorithms, mainly based on “reinforcement learning”, is that a slowly varying environment is assumed and particular access modes, for example C[ode]D[ivision]M[ultiple]A[ccess], is used. These algorithms allow getting a feedback by examining continuously the interference generated by the other nodes.
With this technique in fact nodes are allowed to transmit at the same time, using different codes; in this way if a node increases its transmission power, this creates interference to other nodes, which in order to keep on their own communication increase also their transmission power. In this way the first node is able to see the effect of its first increase in power, and can use this information in the next calculation of its transmission power.
However, these systems do not work in high mobility environments where the channel characteristics vary continuously and unpredictably. Moreover, with typical wireless LAN access modes, for instance C[arrier]S[ense]M[ultiple]A[ccess] type, the examination of the interference does not represent a fast and reliable feedback of the actions operated.
Exemplary systems matching the above description are further disclosed
Various transmission modes to optimize the overall system throughput are well known and implemented in the norm IEEE 802.11 WLAN of the Institute of Electrical and Electronic Engineers. These modes vary the modulation depending on the measured bit error rate. The better the quality of a connection is estimated to be, the higher the bit rate can be chosen.
In general, one of the primary objectives of a wireless local danger warning system is to warn as many nodes, in particular as many drivers, as possible whose life may be endangered for example by some road hazard. The use of existing wireless LAN technologies is attractive because well-tested products are commercially available and supported by the market. However, some functionality is to be added to the system to adapt the performance characteristics to road scenarios.
In prior art article “Distributed Power Control for Reliable Broadcast in Inter-Vehicle Communication System” by Marco Ruffini and Hans-Jürgen Reumerman, VANET 2004 (workshop within MOBICOM 2004 conference), Philadelphia, Pa., USA, Sep. 26 to Oct. 1, 2004, it is proposed to lower the bandwidth consumption by means of regulating the transmission power of broadcast messages without altering the reachability performances.
This mechanism allows efficient dissemination of messages under high and low density traffic situations but the proposed power control concept is only applicable to broadcast mode. In order to reuse the wireless local danger warning system for non-safety related applications, and thereby facilitate the market introduction, a peer-to-peer unicast transmission mode is necessary.
One of the technical challenges in peer-to-peer unicast mode is the trade-off between transmission rate and transmission power, which is required to optimize the network resources. On one side in fact a higher transmission rate increases the network throughput but on the other side it also requires higher transmission power, which increases the interference with other neighbours. Moreover, since broadcast and unicast messages coexist in the same system it becomes compulsory to apply transmit power rules to both message types, otherwise the messages of one type will overpower the others, which will give an unwanted prioritisation to one type of messages.
In wireless LAN networks the ambiguity in the choice of data rate and of transmission power has not yet been resolved.
Starting from the disadvantages and shortcomings as described above and taking the prior art as discussed into account, an object of the present invention is to further develop a controller unit of the kind as described in the technical field, a communication device of the kind as described in the technical field, a communication system of the kind as described in the technical field as well as a method with corresponding communication protocol of the kind as described in the technical field in such way that interference of messages transmitted between and among the mobile nodes is minimized and the overall local network throughput is maximized.
The object of the present invention is achieved by a controller unit comprising the features of claim 1, by a communication device comprising the features of claim 5, by a communication system comprising the features of claim 7, by a communication protocol comprising the features of claim 9 as well as by a method comprising the features of claim 10. Advantageous embodiments and expedient improvements of the present invention are disclosed in the respective dependent claims.
The present invention is principally based on the idea of a safety system for inter-vehicle communication with modulation and power control optimization; in this context, a data rate/transmission power decision algorithm is provided to adapt the data rate and the transmission power on a per packet basis, by gathering and processing information received from neighbouring nodes. In a fully distributed way this system reduces node interference and increases the overall network throughput.
Based on the estimated transmission time, on the averaged neighbour path loss, on the number of nodes interfered and in the probable time wasted for retransmission, a pricing function is calculated for a range of data rate and power margin values. The values of data rate and transmission power resulting in the lowest price is used to transmit the packet.
The present invention solves the ambiguity of choice of data rate and of transmission power, which is typical of wireless LAN networks, in particular by providing a standard defining a formal mechanism of selection of data rate and of transmission power, and is implementable in a completely distributed fashion.
In this context, the present invention is related to the field of power controlled safety systems and methods, in particular to the article “Distributed Power Control for Reliable Broadcast in Inter-Vehicle Communication System” by Marco Ruffini and Hans-Jürgen Reumerman, VANET 2004 (workshop within MOBICOM 2004 conference), Philadelphia, Pa., USA, Sep. 26 to Oct. 1, 2004, where a system for distributing warning messages (using only broadcast messages) among a group of vehicles is described. However, in contrast thereto, the present invention is not restricted to broadcast messages.
According to a preferred embodiment of the present invention the basic functionality of existing wireless local area network (WLAN) systems and methods is exploited and some modifications and improvements are applied to adapt the existing W[ireless]L[ocal]A[rea]N[etwork] systems to the distributed and highly mobile environment of inter-vehicle communication.
According to a particularly inventive refinement, the system as well as the method according to the present invention can generate different types of messages and is able to transmit these different types of messages choosing both transmitting power and data rate. The choice is made in a way minimizing the interference between the neighbours; this technical measure corresponds with the aim of the present system to maximize the overall local network throughput.
Preferably, the system as well as the method of the present invention are designed to decide over which pathloss values their average should be calculated.
According to an advantageous embodiment of the present invention the system and the method use information of pathloss received by neighbouring nodes to adapt the data rate and the transmission power of transmitted packets.
In particular, the system and/or the method may use a price function based on for example
According to an expedient embodiment of the invention the present system as well as the present method associate the minimum price function to the minimum interfering mode, where a mode is referred to as a pair of data rate and of transmit power; in particular, the system as well as the method according to the present invention automatically eliminate the transmitting modes being not compatible with the maximum transmission power.
According to a particularly inventive refinement of the present invention, the system as well as the method according to the present invention retransmit the packet when no acknowledgement signal is received, by recalculating the price function, and favouring the use of a higher power margin.
In general, the present invention, in particular the communication device as described above can be applied and installed in every vehicle moving on road. Said communication device can constitute by itself a complete structure to achieve wireless local danger warning, with the ability of self-adaptation to different circumstances and scenarios.
Moreover, said communication device can also be included as a part of a more complex protocol stack as for example a protocol being designed for retransmitting the message, in particular packet, when no acknowledgement is received, by recalculating the price function, and favouring the use of a higher power margin. For example a general protocol can embody the present invention to solve the general problem of correct choice between data rate and transmission power, independently of the purpose of the communication system and of the data transmitted.
In particular, the present invention finally relates to the use of at least one controller unit as described above and/or of at least one communication device as described above and/or of at least one communication system as described above and/or of at least one communication protocol as described above and/or of the method as described above for at least one wireless ad hoc network, in particular for at least one sensor network or for wireless local danger warning with the ability of self-adaptation to different circumstances and scenarios, for example for car-to-car communication, wherein cars interact cooperatively and distribute for example warning messages, especially for accident-free driving, for instance
Finally, the present invention can also be used for transmitting general data messages to support safety-oriented, telematics-oriented and/or entertainment-oriented applications.
As already discussed above, there are several options to embody as well as to improve the teaching of the present invention in an advantageous manner. To this aim, reference is made to the claims respectively dependent on claim 1, on claim 5, on claim 7, and on claim 10; further improvements, features and advantages of the present invention are explained below in more detail with reference to preferred embodiments by way of example and to the accompanying drawings where
The same reference numerals are used for corresponding features or parts in
In order to avoid unnecessary repetitions, the following description regarding the embodiments, characteristics and advantages of the present invention relates (unless stated otherwise)
Basically, a concept of transmission rate/power decision pertaining to a system 200 (cf.
In correspondence thereto, the algorithm is developed for communication among moving vehicles but can also be embedded in every communication protocol making use of a shared wireless medium.
The general system architecture of the communication device 100 being assigned to the communication system 200 according to the present invention is depicted in
The communication device 100 as shown in
The communication device 100 comprises
The central data processing unit 40, 40′ is configured for calculating the transmitting power and the data rate for sending the messages 22 by processing at least part of the arriving message, in particular by processing information regarding the neighbouring cars 12, 14, 16.
The receiver unit 30 is connected
The receiving/transmitting antenna 23 is assigned to the transmission unit 20 as well as to the receiver unit 30.
For receiving signals
Moreover, the central data processing unit 40, 40′ is connected with a danger sensing unit 90 being designed for sensing one or more subjects being relevant, in particular dangerous, for the considered car 10 and/or for the neighbouring cars 12, 14, 16.
To be supplied with the speed of the respective car 10, the central data processing unit 40, 40′ is connected with a car bus interface 70. Said car bus interface 70 supplies a car bus intra-vehicle system 72 with signals 702 being sent from the car bus interface 70 to the car bus intra-vehicle system 72.
Moreover, the communication device 100, 100′ comprises a display unit 80 displaying messages, in particular the arriving messages, for example the data messages. Said display unit 80 again is connected to the central data processing unit 40, 40′.
Each vehicle 10, 12, 14, 16 equipped with the communication device 100 described in
In the following table, the specification of the neighbour table 410 is depicted, wherein the second column depicts the respective current position of the vehicle 10, 12, 14, 16 and the third column depicts the respective heading direction or moving direction of the vehicle 10, 12, 14, 16:
Additionally other values can be added in the neighbour table 410; for example instead of storing only the history of path loss values, the history of pairs (x=distance; y=path loss) can be stored as depicted in
The hello messages are transmitted in broadcast mode such that every node 10, 12, 14, 16 able to decode the hello message can create an entry in its neighbour table 410 and update the information every time a new hello message from the same node 10, 12, 14, 16 is received.
When no hello message is received from a given node or neighbouring node 12, 14, 16 for a certain amount of time, which can for example be fixed by defining the value of the parameter “max_time” in the system 200, the entry relative to that neighbouring node 12, 14, 16 is deleted from the neighbour table 410 of the reference node 10.
As depicted in
Said receiver interface 430 is connected with a message analysing unit 450 for
To this aim, the message analyzing unit 450 is connected to the receiver interface 430 as well as to the neighbour table 410 as well as to a data managing unit 490, 490′, in particular to a data processing unit or data processor 492, 492′, and provided with the receiving power 504 as calculated by the power estimating unit 50 (cf.
The data managing unit 490, 490′ further comprises
The data managing unit 490, 490′ can be provided with
The data managing unit 490, 490′ is designed for sending a signal 804, 804′ from the central data processing unit 40, 40′, in particular from the message analyzing unit 450 and/or from the data processing unit 492, 492′, to the display unit 80.
The decision unit 482, 482′ is the core of the communication device 100. In said decision unit 482, 482′ a set of rules are defined, which regulate
For transmitting the message 22 being generated by the data message generating unit 460, 460′ or by the hello message generating unit 470, 470′ to the transmission unit 20, the relay central data processing unit 40, 40′ comprises a transmission interface 420 being connected to the decision unit 482, 482′ and being designed for transmitting at least one signal 204 to the transmission unit 20. The decision unit 482, 482′ in turn is connected with the neighbour table 410.
Thus,
Hello messages can be transmitted at variable or preferably maximum power, and together with data messages being processed in the decision unit 482, 482′. This decision unit 482, 482′ uses the information collected in the neighbour table 410 and runs an algorithm determining the best rate and transmission power for this message transmission by following the principles as described by the pricing function (cf. step [ii.d.1] below).
Thereupon, all the messages 22 pass through the transmission interface 420, which is used to adapt the central data processing unit 40, 40′ to the different transmission protocols being usable by the transmitter unit or transmission unit 20 (cf.
Messages 22 incoming from the receiver unit 30 (cf.
Then, messages 22 are passed from the receiver interface 430 to the message analysing unit 450 for deciding if the message 22 arrived is a general data message or a hello message.
In case of hello message, an entry is created (or updated) in the neighbour table 410, with the information provided by the hello message plus the power displayed by the power estimating unit 50; this is inserted into the head of the array of path loss values.
In case the incoming message 22 is a general data message, the central data processing unit 40, 40′ sends a copy to the data processor 492, 492′ inside the data manager 490, 490′ which will process the data message and decide if the income general data message is relevant enough to be displayed at the display unit 80 (cf.
The reference vehicle 10, comprising a communication system 100 as described in
As depicted in
In a first step [i.a], the hello messages as broadcasted by the neighbouring cars 12, 14, 16 are received.
In a second step [i.b],
In this context, “path loss” is the attenuation of the transmission power, in particular of the electromagnetic wave strength, of the message or signal between the time it leaves the transmission unit 20, for example of a neighbouring car 12, 14, 16, and the time it arrives at the receiving unit 30, for example of the reference vehicle 10. The quality of a channel of the communication system 200 depends on the path loss.
In a third step [i.c], the instant path loss for each vehicle is stored in the neighbouring table 410, for example
On receipt of a message 22, the message analyzing unit 450 evaluates if it is a hello message or a general data message:
In the first case, the message analyzing unit 450 uses the hello message to update the neighbour table 410 (cf. neighbour table above) in step [i.f], as will be explained below.
In the latter case, i.e. if the receipt message 22 is a data message the message analyzing unit 450 sends this data message to the data processing unit 492, 492′ to evaluate if the message has to be sent to the display 80 and if the data message has to be further processed and eventually supplied to the data generating unit 460, 460′.
In a next step [i.e], the average path loss and the path loss variance per neighbour 12, 14, 16 are calculated.
The neighbour table 410 includes for each entry the same information contained in the correspondent hello message, plus an array of recorded path loss values (cf. neighbour table above), the most recent of which is provided by the power estimating unit 50 (cf.
The averaged path loss value is obtained by averaging a variable number of instant path loss values, which are calculated by subtracting the value received by the power estimator 50 (cf.
The path loss variance is instead the variance of the instant path loss values considered in the averaged path loss, calculated following the usual formula
where μ represents the averaged path loss.
According to the present invention, in the calculation only the values xi where the path loss is higher than the average value/are included as clarified below (cf.
It is important that the number of values over which the average is calculated is variable, and it has to be chosen considering estimations made about the channel characteristics and hello messages repetition rate.
Between steps [i.c.1], [i.c.2], [i.c.3] and step [i.e], a step [i.d] of calculating the general path loss characteristics plc (cf.
The path loss average is computed from a number of path loss values regularly measured by comparing transmit power and receive power from hello messages delivered by neighbouring vehicles 12, 14, 16. Each vehicle 10, 12, 14, 16 averages the path loss value to be independent of the fast fading effect of the transmission channel, where fading refers to the signal deterioration due to multiple reflections at stationary objects and/or at moving objects.
If the frequency of path loss measurements is too low, then the environment may have changed a lot and consequently the fading effect is not sufficiently considered, while if the frequency is too high then the free space loss effect dominates and deteriorates the quality of the averaging process.
Therefore, the frequency of path loss measurements, closely depending on the frequency of hello messages sent or received, can advantageously be varied considering an estimated free space loss for the channel which can be derived and extrapolated from the information contained in the hello messages. The extrapolation can be done by means of existing methods, as for instance minimum square difference.
When averaging the path loss derived from the hello messages transmitted for example every hundred milliseconds from the same vehicle 10, 12, 14, 16, it can be assumed that the environment of the vehicle remains stable and the fading effect can be neglected.
Taking into consideration that regularly hello messages from more than one vehicle are received, and that those vehicles move very slowly compared to the hello message rate, a path loss model can be approximated and a graph on received power (=variable y in
As depicted in
If for example the values of path loss change are within one decibel to two decibel for a distance between 120 meters and 130 meters, then all the path loss values coming from the same neighbour within this range can contribute to the calculation of the average path loss for this neighbour.
Referring to
As further depicted in
First, the interference time per available bit rate is calculated in a step [ii.a].
Every time a node, for example the reference vehicle 10, wants to transmit a hello message and/or a data message, for example a packet of messages 22, the node calculates the estimated time durations of the transmission for every data rate available because the transmission time depends on the data rate used.
The estimated time duration represents the effective time used to send the packet; therefore the estimated time duration has to include preambles and all the various system overhead; for example, in the IEEE 802.11 case also the D[istributed Coordination]I[nter]F[rame]S[pacing] parameter has to be considered.
These values are stored in variables named Tmod, where “mod” indicates the type of modulation or data rate considered. The formula for transmission time calculation is available in prior art document “802.11 Efficiency Analysis” by F. Dalmases, PFL-Aachen Technical Note 9/2002.
In the step [ii.a] of calculating the interference time per available bit rate also the time spent for the eventual use of acknowledgments (ACK) or other transmission related system functionalities can be included, for instance R[eady]T[o]S[end]/C[lear]T[o]S[end] mechanism as disclosed in “The International Standard ISO/IEC 8802-11, Part 11: Wireless LAN Medium Access Control (MAC) and Physical Layer (PHY) Specifications”, 1999(E) ANSI/IEEE Standard 802.11.
The following table explains the selection of the data rate and of the transmission power as described above, wherein it is assumed that the message 22 to be transmitted comprises a packet length of 120 bytes. The column “time to transmit the packet Tmod(μs)” considers the time to send the message 22 or packet and the time to receive the corresponding acknowledgment:
The step [ii.b.1] of determining the path loss to the receiver 30 is to calculate the power required in order to have the transmitted packet or message 22 correctly decoded at the receiver unit 30 (cf.
Every different modulation has a fixed sensitivity value Smod indicating the minimum power needed to decode the message 22. By summing each of these sensitivity values Smod to the averaged path loss Plavg, a set of values Pmod(=Smod+Plavg) is obtained wherein the values Pmod indicate, for each modulation, the power at which the node has to transmit its message in order to reach the receiver unit 30 with the correct power.
The following table explains the step [ii.b.1] of determining the path loss to the receiver 30 as described above, wherein it is supposed that a message 22 with an average path loss value Plavg of 102 decibel is to be transmitted to the second node. Moreover, it is supposed that in step [ii.b.2] of calculating the required transmission power per available bitrate the maximum transmission power (E[ffective]I[sotropic]R[adiated]P[ower]) is thirty decibel:
The figures 37 dBm (=decibel milliwatt), 36 dBm, 32 dBm in the first three lines represent values going over the maximum transmission power of the transmitter, for instance over thirty decibel milliwatt.
After
for every modulation different or various power margin values have to be considered and applied in step [ii.c.1], which are added to the value Pmod calculated. The power margin is used to increase the transmission power; in fact the Pmod values are calculated on the basis of the average path loss.
Since, as stated above, the actual path loss is a probabilistic value, increasing the transmission power (by increasing the margin) also increases the probability of correct message reception. The different margin values are stored in variables named Mn. In this context, it has to be taken into account that increasing transmitted power also increases the communication interference experienced by the other nodes.
For each couple of values {Pmod, Mn} the sum Pmod,ntot=Pmod+Mn is calculated, and to this number Pmod,ntot the number of neighbours Nmod,n is associated which will be interfered when using this transmission power.
In this context, a node is considered to be interfered when it detects a minimum power level for which it considers the transmission medium as busy. In the case of IEEE 802.11 this corresponds to the threshold thcaa of the channel assessment avoidance. In this way, the value of the path loss within which the nodes are interfered can be calculated by the formula Imod,n=Pmod,ntot−thcaa.
The value Nmod,n is calculated from the neighbour table 410 representing the number of neighbours 12, 14, 16 the average path loss value of which is less than Imod,n.
In the following,
In the following table, these steps [a], [b], [c] as described above are depicted, wherein
The following table depicts a detail of the neighbour table 410:
The following table depicts the values Nmod,n, i.e. the number of nodes interfered, depending on the modulation and on the margin used:
At this point all the values calculated are included in a pricing function pricemod,n considering all the benefits as well as all the disadvantages of the different modulations and transmission powers calculated. The values of modulation and of transmission power minimizing the pricing function pricemod,n are assumed to be the best values from the point of view of optimisation of the local network performances.
The pricing function is pricemod,n=Tmod·Nmod,n+Tmod,nre-tx·Nmod,n represents the time Tmod a transmission occupies the wireless medium, multiplied by the number of nodes Nmod,n interfered. In this formula for the pricing function pricemod,n, it is also taken into account that increasing the margin increases the probabilities of correct reception, while a corrupted messages implies a wasted time for retransmission; thus, the term Tmod,nre-tx indicates the time wasted to retransmit the message 22 in case this is not correctly received; this value has a probabilistic nature and decreases as the margin increases.
The step [ii.d.1] of minimizing the interference time with neighbours 12, 14, 16 is described in more detail as follows:
Considering the equation pricemod,n=Tmod·Nmod,n+Tmod,nre-tx·Nmod,n for the pricing function pricemod,n,
The communication device 100, in particular the decision unit 482, 482′, calculates this expression for all the previously considered values of the data rate and of the power margin, using the so calculated values of Tmod, Nmod,n, Tmod,nre-tx.
Once all the various pricing functions pricemod,n are calculated, the communication device 100, in particular the decision unit 482, 482′ extracts the smallest value of pricemod,n to which the data rate “mod” and the transmission power margin “n” has been associated.
At this point the communication device 100, in particular the transmission unit 20, transmits the message 22 using the data rate “mod” and the power margin “m”.
The value Tmod,nre-tx of the time wasted to retransmit the message 22 depends on the variance of the path loss: in fact higher variance implies that higher margin has to be used to assure a certain bit error rate. This value Tmod,nre-tx can be calculated as Tmod,nre-tx=Tmod·probn, where probn indicates the probability that a message 22 transmitted with the margin “n” is not correctly decoded by the receiver unit 30.
In the following, the step [ii.d.2] of selecting the resulting power/bit rate value is described in more detail:
The term or value Tmod,nre-tx(, wherein Tmod is the same value as in the pricing formula while probe is the probability that the message 22 is not received) indicates the time wasted due to the fact that the message 22 is not correctly received. In this case, in fact the communication device 100, in particular the transmission unit 20, tries to re-transmit the message 22.
In this context, the term probn is probabilistic and depends on the margin “n” used to transmit; in order to calculate probn the probability of the received power being lower than the power needed to correctly decode the message 22 needs to be understood. This is equal to the probability that the instant path loss will overcome the average path loss of a value higher than the power margin chosen. So the calculation of the probability that the fast fading overcomes the chosen power margin is interesting.
The random fading is approximated with a Gaussian random variable, wherein the probability distribution function of the Gaussian random variable is the well-known expression:
The value of σ is known from the path loss variance; the value of μ is known from the average path loss.
By integrating the formula for f(x), the probability of random fading being lower than a given value of margin “n” can be obtained:
Since the probability of the fading being higher than the margin is required the term 1−F(n) is considered. In this context, tabled values of F(n) are available in standard mathematics tables related to the function erf(x).
Finally; the value used in the pricing formula is probn=1−F(n).
After the step [ii.d.2] of selecting the resulting power/bit rate value, the message 22 can be sent in the final step [ii.e] by the transmitting unit 20.
The embodiment of the present invention as described above may advantageously further comprise one or more of the following details:
The path loss variance considered in the calculation (cf. steps [i.e] and the following above) is derived from the history of the path loss values stored in the neighbour table 410 and can be differentiated for every node 10, 12, 14, 16 or be averaged over a certain number of nodes or even over all nodes 10, 12, 14, 16.
The priority of the messages 22 also advantageously can be considered when calculating the transmission power; messages 22 with higher priority in fact can be transmitted with higher margin to increase the probability of being correctly decoded at the first transmission attempt. This condition can be included in the pricing mechanism (cf. step [ii.d.1] above) considering that a node 10, 12, 14, 16 is willed to “pay a higher price” to give higher reliability to a message 22 with higher priority.
Also the problem of packet collision or message collision can be taken into consideration in the pricing formula pricemod,n=Tmod·Nmod,n+Tmod,nre-tx·Nmod,n (cf. step [ii.d.1] above). In fact, transmitting at higher power, beyond preventing other stations or nodes from using the channel can also increase the probability to create a packet collision at a receiver node, in particular at the receiving vehicle. So, a further term can be included in the pricing function pricemod,n=Tmod·Nmod,n+Tmod,nre-tx·Nmod,n this further term penalizing a further increasing in the transmission power.
When an expected acknowledgement (ACK) is not received, the communication device 100 can attempt to retransmit the message 22, recalculating again the transmission power and the data rate from the previous formula (cf. steps [ii.b.2] and the following above) but considering that a higher price can be paid by the communication device 100, similar as explained for priority handling above.
In this way, a higher margin can be chosen that will increase the probability of correct reception. This is considered in the pricing function pricemod,n=Tmod·Nmod,n+Tmod,nre-tx·Nmod,n by increasing the value of the term representing the probability probn of packet error.
If no acknowledgement (ACK) is received after a fixed number of trials, the communication device 100 stops attempting retransmission and sends a message back to the data manager 490, 490′ to inform that message 22 cannot be delivered.
The present method tends to solve two different aspects of the interference problem; in fact, the interference towards a vehicle can be seen in two different ways. On one side the vehicle requiring to send a message 22 is prevented from transmitting because it senses the medium busy; on the other side, a vehicle receiving a message 22 can be prevented to decode it correctly because another message 22 interferes with this message 22.
The first case is largely considered along the present invention; the second case is known as hidden node problem. In this context, it may be noticed that the pricing function takes into account both situations at the same time: the capability of diminishing the number of neighbour nodes, in particular neighbour vehicles 12, 14, 16, within the zone of interference in fact is advantageous for the two aspects of the interference problem.
The communication device 100 can also be included as a part of a more complex protocol stack for multi-purpose communication system 200.
The communication of a reference vehicle 10 with a first neighbour vehicle 12 being within an inner circle c1 at a high data rate requires a transmission power interfering with too many third neighbour vehicles 16 at the border area, i.e. being arranged between a middle circle c2 and an outer circle c3. Thus, the pricing function (cf. step [ii.d.1] above) favours a transmission at a lower data rate and at a lower power with lower interference. The resulting interference corresponds to the middle circle c2.
In this case, not many neighbours 12, 14, 16 are within the range of interference of the transmission at high data rate corresponding to the outer circle c3. Thus, the communication of the reference vehicle 10 with the first neighbour 12 within the inner circle c1 at a high data rate can be achieved without disturbing other neighbours 14, 16. Hence, the pricing function favours a transmission at high data rate and at high power in order to reduce the time occupancy of the channel because choosing lower data rate (<--> middle circle c2) will not reduce interference with other vehicles 14, 16.
The central processing unit 40′ is extended to include the functionalities of a communication device 100 being designed for calculating the transmission power in broadcast communication by processing information received from neighbouring nodes 12, 14, 16, in particular for calculating the path loss of every neighbouring node 12, 14, 16 by using the difference between the power transmitted value and the power at which the message 22 is received (cf. prior art article “Distributed Power Control for Reliable Broadcast in Inter-Vehicle Communication System” by Marco Ruffini and Hans-Jürgen Reumerman, VANET 2004 (workshop within MOBICOM 2004 conference), Philadelphia, Pa., USA, Sep. 26 to Oct. 1, 2004).
The central data processing unit 40′ comprises the same components as described in
The retransmission controlling unit 440′ is connected with the neighbour table 410, with the message analyzing unit 450 and with the decision unit 482′ being assigned to the power control subsystem 480′.
Said power control subsystem 480′ is designed for sorting the information regarding the neighbouring nodes 12, 14, 16 in the neighbour table 410 according to increasing path loss calculation values.
The following table specifies the details of the table 410 of the neighbours 12, 14, 16 and displays the grouping of the neighbour cars 12, 14, 16 in different path loss intervals or classes, as actuated by the power control subsystem 480′ (cf.
The power control subsystem 480′ is connected with the transmission interface 420, with the neighbouring table 410, and with the data managing unit 490′ comprising the data message generating unit 460′ including a warning message generating unit being designed for providing the power control subsystem 480′ with one or more warning messages.
Moreover, the data managing unit 490′ comprises
Thus, according to the central processing unit 40′ messages can be generated
Finally, some typical scenarios are given where the communication system 200 can operate to deliver warning dissemination.
The communication system 200 is relevant for car-to-car communication where sensor-equipped cars 10, 12, 14, 16 interact cooperatively to avoid collisions. For example, car-to-car communication is considered crucial for intersection collision avoidance, in particular to avoid collisions when cars 12 are entering an intersection that should be kept free for instance for a fire truck 10 (cf.
Likewise, the communication system 200 according to the present invention can be used for cooperative interaction of cars 10, 12, 14, 16 and for distributing messages 22, in particular warning messages, especially
| Number | Date | Country | Kind |
|---|---|---|---|
| 04103518.9 | Jul 2004 | EP | regional |
| 04106841.2 | Dec 2004 | EP | regional |
| 05102498.2 | Mar 2005 | EP | regional |
| Filing Document | Filing Date | Country | Kind | 371c Date |
|---|---|---|---|---|
| PCT/IB2005/052476 | 7/22/2005 | WO | 00 | 1/18/2007 |
