DE4012904A1 - Adaptive routing and media access control for mobile communications - using decentralised control via information tables for all stations or those within relevant area - Google Patents

Abstract

The adaptive routine and media access control uses information obtd. from a digitised table and relating to all the stations within the mobile communication network, or only to those within a relevant area, the cluster within which the mobile station is located being monitored, to evaluate the cluster which will be reached at a given time point. The tables are updated in dependence on the new stations entering each area, with dynamic addressing of the stations, each station identification followed by a cluster identification, the routing and medium selection utilising the optimum path and the preferred medium without risk of conflict. ADVANTAGE - Rapid table updating.

Description

The invention relates to a knowledge-based method according to the Preamble of claim 1.

In a mobile communication network, both voice services as well as non-voice services are offered. The Signal transmission can be analog or digital. The Access to the transmission medium is through appropriate Access procedure regulated. If a direct Signal transmission between communicating stations physical, technical or functional reasons not communication is possible via other (mobile or fixed) stations take place, which (in this case) as an operator (Relay) function. Communication between stations can through line switching or packet switching and connection-oriented for packet switching or be datagram-oriented. The mobile communication network can either be a subnet of a larger communication network or through certain (fixed) stations with other public and private communication networks. Depends on the Speed and direction of movement of the mobile stations there is a more or less rapid change in the Network topology, which allows for communication with mobile Stations via mobile or fixed stations that act as brokers (Relay) serve, adaptive route selection procedure required  are: Each station that acts as an operator chooses one optimal way out on the optimization of one or more relevant criteria (e.g. minimizing the number required switching stations along the selected one Way, cost minimization, runtime minimization, maximization of Throughput) based on the following requirements should be:

• - The network topology on which the route selection is based the current network topology should correspond to non-optimal routes or misdirected data packets avoid;
• - The volume and number of network control information which are broadcast by the stations and within the Relevance zone to be disseminated to give an exact picture of the To achieve the current network topology should be as low as possible being held.

Several adaptive methods for route selection (routing) for are mobile packet radio networks with a large number of stations known. Because of the mobility of the stations Routing procedures become more complex for such networks usually hierarchical structures and corresponding Routing strategies suggested. Such suggestions can be found u. a. in Proceedings of Globecom '84, IEEE CH2064-4 / 84 / 12:00 0468, pages 468-471 and Proceedings of Infocom '85, IEEE CH2153-5 / 85/0000/0292, pages 292-302. As reasons for that hierarchical structure is explicitly included in the latter Compromise between

• - scope of the routing tables,
• - Number of update messages (English "updates") and
• - The speed at which these updates spread across the network called.

The main problems of adaptive routing, namely the number of update messages in the network Management of the routing tables need to be spread, and the speed at which these messages appear on the network will be spread, tackled, but in the  Contrast to the knowledge-based adaptive route selection process of the present invention, which is completely dynamic the known routing procedures network topology changes then spread when they have occurred, so, accordingly the mobility of the stations and the speed with which this network status information spread that Information may already be out of date (i.e. no longer the reflect current network topology) if it continues at distant stations arrive, making it less than optimal Routes and misdirected data packets can come. In addition, every network topology change triggers one appropriate status message, resulting in a significant Network load.

For the distribution of network status information to everyone Stations within a network can use procedures like that in the Proceedings of 39th IEEE Vehicular Conference, San Francisco, May 1989, pages 793-798 will. In this broadcast process, by hop-wise Multicasting the sending of a packet to a station that forwarded a similar packet via another station is sought to avoid, which leads to a lesser Network load than comes with pure flooding.

In contrast to the mobile packet radio networks where usually every mobile station also as an intermediary (relay) can act to communicate to other stations enable, takes place in the cellular mobile radio networks (C network, European mobile communication network according to CEPT / GSM (D network), etc.) only communicate with mobile stations Fixed stations that act as relays (for the transition to wired Networks). This simplifies the required Routing procedures considerable, but also here adaptive procedures are applied to the mobile subscriber to reach. In the Proceedings of DCRC, Hagen, Westphalia, FRG, October 12-14, 1988, pages 4b / 1-4b / 12 and pages 4c / 1- 4c / 11 will be for the GSM Public Land Mobile Network Methods known (European Mobile Communication Network)  (English "Location Updating"). It must go through any mobile station that moves from a "location area" into a neighboring "location area", one Update message to the corresponding "Home Location Register "to announce which new "Visited Location Register" now for the mobile station contains significant communication parameters. Opposite this Event-driven update notification can also periodic updates are sent to reachability announce a mobile station to the network and To correct incorrect entries in the "Location Registers".

Although the update messages are not sent to all stations need to be disseminated, as the number of participants increases, high mobility, small "location areas" and great distance to the "Home Location Register" also by this procedure not inconsiderable network load.

For controlling access to the transmission medium (English "Media Access Control") are both asynchronous as well synchronous method used. The division of the transfer mediums between the stations with communication needs within the transmission / reception range is done by frequency Multiplex (English "frequency division multiplex"), time Multiplex (English "time division multiplex") or code Multiplex (English "code division multiplex"). Sending two Stations simultaneously on the same frequency band, within of the same time slot (English "slot") or with the same Code word, so it can be at receiving stations within the Interference ranges come to collisions due to interference, thereby the correct receipt of the sent messages becomes impossible. Outside the interference range, a after channel formed by one of the above methods be reused. Most access problems in one mobile environment arise from the limited knowledge of Stations about the location, the mobility behavior as well as the communication needs of other stations are within their transmission or interference range.  

Channel access protocols are known which have been developed so the stations at least have information about their 2-hop Own environment. A corresponding protocol can be found in the Proceedings of 9th ICCC, Tel Aviv, Oct./Nov. 1988, pages 83-88; an improved access protocol based on this is published in the IT reports 205, Springer Verlag, Berlin, Feb. 1989, pages 280-293. Both Logs provide information about the current Channel occupancy available to avoid channel access conflicts to reduce. This information is given by one of the number of slots in transmitted a bitmap corresponding to a frame, which in each Slot of a frame (240 slots / frame, 20 frames / s) sent out becomes. This bitmap process therefore causes a very high one Overhead.

The knowledge-based process for controlling access to Transmission medium of the present invention enables also information about the current channel assignment, however, in contrast to the known methods, not only for a 2-hop environment, but for an n-hop environment (with any size n). In addition, this is not done by frequent Send out within a second, as with the known ones Procedure, but only at time intervals which are approximately in the Minute range is reached, with the update notification of the knowledge-based method according to the invention in is usually even shorter than a single (!) bitmap, and nevertheless current information about numerous important Network status parameters (network topology, channel assignment, Transmission power, etc.) enables the corresponding knowledge-based tables managed decentrally, but quasi be updated synchronously. This can do the possible The occurrence of access conflicts is calculated in advance and therefore Collisions are avoided.

The invention is based, in one Mobile communication network, in particular also in such a a variety of high mobility stations, suitable  Route selection, addressing and medium access procedures for the Communication between any stations is available too put so that despite a rapidly changing network topology optimal routes (routes) are selected, Access conflicts when accessing the transmission medium be avoided and the data volume of network status information to achieve these goals is as low as possible is held.

This task is carried out in a generic communication network solved by the characterizing features of claim 1. The further claims contain advantageous developments of Invention.

Significant advantages of the knowledge-based according to the invention Procedure for adaptive route selection and control of access regarding the transmission medium compared to known methods are:

• - Station status or update messages (the important Information to perform the functions of network and Switching level included), must (if not to switch the mobility behavior of the stations) is relatively rare sent out and further distributed within the relevance zone be, resulting in a lower load on the network and thus a u. U. significant increase in throughput of User data results.
• - Nevertheless, all stations (also with communication networks with a large number of stations with high mobility) constantly accurate and extensive information about the current network topology.
• - It is even a prediction about the future network topology possible.
• - Because of the precise and extensive information about the current and even future network topology can be one efficient knowledge-based adaptive route selections take place, being because of the decentralized and yet at all stations  quasi synchronous update of the knowledge-based tables, suboptimal routes, loops (English "loops") and misdirected data packets largely can be avoided.
• - In contrast to known methods, the stations have also current for distant areas (clusters) Information about the network topology.
• - Because of the property of the invention knowledge based routing that the availability of Routes can be calculated in advance, it is possible Only send data packets at the right time if reaching the target station is also very likely is, which results in a significant reduction in Network load may result.
• - Through accurate and extensive information about the current and even future network topology can be the available capacity of the transmission medium is also optimal use and largely avoid access conflicts without additional resources used or more permanent, extensive Exchange of information between neighboring stations (such as in part in known methods) would have to be carried out.
• - The dynamic, knowledge-based addressing is also the integration of stations possible that do not invent this knowledge-based processes according to the invention, but conventional ones Use routing procedures.

Because of the significant advantages, the invention can be use it practically for every communication network, especially those with a large number of high stations Mobility that move on predetermined traffic routes, that is for communication with stations in road vehicles, Trains, planes, ships, satellites with lower Orbit, missiles and other missiles. As examples be given:

• - Cellular mobile phone network (car phone, C network);
• - GSM Public Land Mobile Network (European Mobile Communications tion network, D network);
• - PROMETHEUS (project of the European automotive industry: Communication between vehicles and with infrastructure);
• - DRIVE (European Community project: Communication between vehicles and with infrastructure);
• - Mobility 2000 (project of the United States of America: Communication between vehicles and with infrastructure);
• - Comparable communication networks in Japan;
• - train radio;
• - aeronautical radio;
• - battlefield communication;
• - Land mobile radio (communication with vehicles via satellite).

The invention is described in detail below.

The generic communication network of the invention has following features:

• - The mobile stations can have their own current Determine position, speed and direction of movement or estimate;
• - The stations can use a Determine accuracy of at least one second or estimate;
• - The communication network is based on a single-level or multi-level hierarchical division into fields (hereinafter referred to as "cluster" or "1-cluster", "2-cluster",..., "n-cluster") as shown in Fig. 1 , which is known to the stations;
• - This division can be geographical or logical (e.g. based on a traffic network) aspects respectively;
• - When dimensioning the clusters Transmit / receive radii of the stations taken into account;  
• - each station continuously manages and updates the for route selection, addressing and medium access required information about all stations in the Area of relevance using knowledge-based tables ("Knowledge-based cluster table", "Neighbor table", "Cluster routing table") so that the cluster in which there is currently a mobile station, so to speak can be determined without delay and also a Prediction of reaching a certain cluster possible is;
• - for the quasi-synchronous update of the by Stations decentrally managed knowledge-based tables, every new station that is added sends the required ones Information to the neighboring stations from these, if necessary, continue within the relevance zone to be spread; by implicit or explicit confirmation through the neighboring stations the new one arrives Station knowledge of their direct neighboring stations, on what they all from a neighboring station to set up their own queries knowledge-based tables for information needed;
• - the new station already has its own turn built up knowledge based tables (meeting of Subnets), the necessary information for the Extension of these tables between the participating stations exchanged and, if necessary, within the Relevance zone further spread over neighboring stations;
• - each station constantly compares the ones it sends out Information required to update by stations decentralized and quasi-synchronously managed knowledge-based Tables serve with the actual values and sends, at Deviation beyond the specified tolerance range (i.e. if the station finds that the distance to that at the time for the station itself in its own cluster table - and due to the method according to the invention also in the cluster Tables of all other stations in the relevance zone - saved cluster is larger than the specified one Tolerance value), as well as independently, when a system-wide time interval since the last Transmission of the current values to all neighboring stations  of these, if necessary, within the relevance zone spread more widely;
• - According to the extension obtained in claim 2 know all or individual stations in certain clusters existing speed limits or can query so that the stations when updating the advanced knowledge-based regarding this information Tables, the expected maximum speed of a station when driving through the cluster take into account what the Sending messages with current values (because of the Speed limit) by the corresponding station is largely avoided;
• - According to the extension obtained in claim 3 can all or individual stations use the mean value and the Standard deviation of intended cruising speeds the stations in a cluster (taking into account the Direction of movement) as well as the mean and the Standard deviation of the actual speeds determine or query and thereby when updating the advanced knowledge-based regarding this information Tables showing the expected speed of a station Drive through the cluster, taking into account the emission of messages with current values by the corresponding Station is largely avoided;
• - The stations are addressed dynamically and contains the identifier of the cluster in addition to the station identifier the station is currently in;
• - for the route selection (English "Routing") those of the stations managed decentrally, almost synchronously updated knowledge-based tables evaluated to the optimal way (route) to a station in a certain Determine clusters;
• - for access to the transmission medium are those of the stations managed decentrally, almost synchronously updated knowledge-based tables evaluated to that Optimal use of the transmission medium and access conflicts avoid.

Cluster division

The communication network is system-wide a single-level or multi-level (n-level) hierarchical division into fields (hereinafter referred to as "cluster" or "1-cluster", "2-cluster",..., "N-cluster") according to Fig . 1 basis. This subdivision can be based on geographical or logical considerations. The length and width of the clusters can be the same or different. The cluster structure is known to all stations. The dimensioning of the clusters should be selected so that with regard to the transmission / reception ranges of the stations and taking into account a definable tolerance factor, every station in a cluster can be reached directly from each station (see Fig. 2).

Dynamic addressing

Each station is identified by a unique station identifier and the Identifier of the cluster in which the station is currently located (according to the knowledge-based cluster tables) characterized:

<Station address <:: = <station identifier << cluster identifier <.

If the in the message header (header) of a data packet included address for a specific station with the (for the same station identifier) in the knowledge-based Cluster table (see below) contains the address matches, this will be in the cluster table designated address as new station address dynamic assigned.

It is through this method of dynamic addressing possible, even stations that are not the knowledge-based process of the invention for adaptive route selection but others Use routing procedures in the communication network too integrate.  

Knowledge-based tables

Each station in the communication network manages and updates the following knowledge-based (digital) tables:

• - Knowledge-based cluster table (dynamically updated Table)
• - Direct neighbor table (contains directly accessible neighbors)
• - Cluster routing table (for knowledge-based adaptive Route selection)

The and contained in these knowledge-based tables updated parameters (shown in square brackets) that current values for each entry are shown below explained in more detail. The unit used and the range of values the parameter depends on the particular application of the Communication network and is therefore suitable for every application to be determined.

Knowledge-based cluster table

<Entry number <: consecutive number of the entry.

<Station identifier << cluster identifier <:

Dynamic address of the station of the respective entry.

The current cluster identifier shows in which cluster (taking into account the specified tolerance values) the station is currently.

<Position <:

This optional parameter gives the current position of the Station at the time of sending the last cluster Update notification and can for more precise determination predicted positions and for checking of the reported parameters can be used.

<Speed <:

This optional parameter gives the current speed station at the time of sending the last cluster update message and can go to more precise determination of predicted positions as well be used to check the reported parameters.

<Acceleration <:

This optional parameter gives the current acceleration station / delay at the time of transmission the last cluster update message and can be used for more precise determination of predicted positions as well be used to check the reported parameters.

<Desired speed <:

This optional parameter gives the desired speed station at the time of sending the last cluster update message and can go to more precise determination of predicted positions as well be used to check the reported parameters.

<Direction of movement <:

Physical or logical direction of movement of the station.

Specially defined values cannot currently be in Mobile station in motion, a mobile station with indefinite (constantly changing) direction of movement or mark a base station. Mobility dependent In such cases, parameters can be disregarded stay or be interpreted differently.

<Airtime <:

Current time when the last cluster updates were sent sation message for those listed above more precise position predictions and parameter checks gene is needed.

It also makes it possible to have signal delays calculate that is important for routing decisions are.

<Next arrival <or <Next x arrival <:

Time at which the station is expected to be the next (or the next cluster boundary in the x direction) reached.

If the current time exceeds the specified time Arrival time, this is by the in <pass interval <(or <Pass-x interval <, see below) increment the specified value and the <cluster identifier <(see above) on the Movement direction corresponding new value set.

<Next y arrival <:

Time at which the station is expected to be the next in cluster limit in y-direction reached.

The increment (using <Pass-y-Inter vall <) and the update of the <cluster identifier < is the same as for <Next x arrival < specified.

This parameter can be omitted if the system-wide Subdivision into fields according to logical criteria was hit, and the stations were only in 4 logi directions parallel to the cluster boundaries  move. In this case, only the for all directions <Next arrival <parameter (see above) is used.

<Next z arrival <:

Time at which the station is expected to be the next in Cluster boundary in z-direction reached.

This parameter is only used in a communication network used according to claim 5.

<Pass interval <or <Pass x interval <:

Time interval after which the reaching of the further cluster boundaries (or further in the x direction cluster boundaries) depending on the current Direction of movement and the current (or intended) Speed of the station is expected.

This value gives the expected time to pass the following cluster and is used to increment the Arrival times at the cluster borders (see above) drawn.

<Pass-y interval <:

Time interval after which the reaching of the further cluster boundaries in the y direction in Dependence of the current direction of movement and the current (or intended) speed of the Station is expected.

This parameter can be omitted if the system-wide Subdivision into fields according to logical criteria was hit, and the stations were only in 4 logical directions parallel to the cluster boundaries move. In this case, only the for all directions Parameter <pass time <(see above) is used.

<Pass-z interval <:

Time interval after which the reaching of the further cluster boundaries in the z direction in Dependence of the current direction of movement and the current (or intended) speed of the Station is expected.

This parameter is only used in a communication network Claim 5 used.

<Validity cluster <:

Identifier of the cluster up to which it is expected that the from values reported to the station will be valid.

This is intended to calculate positions in Clusters that the station will never reach be avoided.

<Relevance zone code <:

Definable code that indicates how far that from one Station reported information is relevant and therefore to be distributed to other stations.

<Function code <:

Code indicating which embodiment of the invention ent speaking the station behaves and what function it thereby fulfilled.

For example, the <function code <could be as follows be constructed:

<Function code <: =
00 <function << variant << main variant <

<Main variant <:
Bit 0.1 with the following values:
0. . . <reserved <
1 . . . Claim 1
2nd . . Claim 2
3rd . . Claim 3

<Variant <:
Bit 2.. . 5 with the following meaning:
bit
2 set. . . Claim 4
3 set. . . Claim 5
4 set. . . Claim 6
5 set. . . Claim 7

<Function <:: =
Bit 6.. . 13 with the following meaning:
bit
6 set. . . Station that currently does not perform any of the functions below, but could do so
7 set. . . Station with relay function
8 set. . . Station with operator function
9 set. . . Information station
10 set. . . Control station
11 set. . . Control station

bit
12, 13. . . Hierarchy level with the values:
0. . . no hierarchy
1 . . . Hierarchy level 1
2nd . . Hierarchy level 2
3rd . . Hierarchy level 3

bit
14, 15. . . <00 <reserved

Examples

<Update time <:

Time when the last update message from one Station was received.

This makes it possible for stations to have the relevance zone have left or been switched off (without this beforehand to announce) after a definable interval delete from the cluster table.

<Frequency identifier <:

This optional parameter (possibly also list of Parameters) for a knowledge-based according to the invention Access to the transmission medium is provided when using Frequency division multiplex (English "frequency division multiplex ") which frequency (channel) (or which Frequencies) is currently occupied by the station (will).

Through the knowledge based method of the invention based possibility of the advance calculation in which Cluster the stations at a specific time the available channel capacity can be expected dynamically and decentrally managed and therefore optimal  be used. The stations can be timely (and change the channel independently if (due to the Mobility) collisions (by using the same Channel through other stations within the relevant Range of interference) are to be expected, whereby conflicts in the Access to the transmission medium should be avoided before they arise.

Because all stations are from other stations know channels occupied within the relevance zone Newly added stations after querying the builders of the knowledge-based tables required (via a dedicated organization channel) one still completely free or considering the own direction of movement only in distant clusters or Select a rarely used channel so that the number necessary channel change is minimized.

<Slot identifier <:

This optional parameter (possibly also list of Parameters) for a knowledge-based according to the invention Access to the transmission medium is provided when using Time multiplex (English "time division multiplex") on, which time slot (English "slot") in a time frame (English "frame") or which time slots through the Station is currently being occupied.

With time division multiplexing, a channel can be created using the same time slot in successive frames be formed. By based on the knowledge-based Method of the invention based possibility Predicting in which cluster particular stations can be expected at a certain time existing channel capacity (i.e. available standing slots within a time frame) dynamic and managed decentrally and thus optimally used. The stations can be on time (and independently) Change channel (slot) in the event of collisions (through use  the same slot in a time frame by multiple Stations within the relevant interference range) are expected, causing conflicts when accessing the Transmission medium be avoided before they arise.

The same applies to new stations added above analogous.

<Code word identifier <:

This optional parameter (possibly also list of Parameters) for a knowledge-based according to the invention Access to the transmission medium is provided when using Code multiplex (English "code division multiplex"), which code word (or which code words) for the spread of the frequency band (s) through the station is currently being used, if not system-wide a fixed code word to spread the over supporting signals is used.

Through the knowledge based method of the invention based possibility of the advance calculation in which Cluster specific stations at a specific time the stations can be expected on the one hand the code word used when sending by other stations use around those broadcast by these stations Receiving messages, but also right timely (and independently) ver Change codeword if collisions occur (by ver the use of the same code word by other stations are to be expected within the relevant interference range) causing conflicts when accessing the transmission medium be avoided before they arise.

The same applies to new stations added above analogous.

<Transmission power <:

This parameter currently gives the through the station used transmission power, whereby the expected Sending / receiving and interference range estimated can be.

<Antenna code <:

Optional code (or code list) that specifies which Antenna characteristics, horizontal and vertical Beam angle, antenna inclination, main beam direction at directional antennas or antenna rotation speed and antenna position at certain times when rotating Antennas, polarization level, etc. currently through the Station used, which, taking into account the specified transmission power, the expected transmission / Reception and interference range in certain directions can be estimated.

<Environment code <:

Optional code (or code list) that specifies in which environment (e.g. open area, hilly country, forest, built-up area) and / or under which weather conditions (e.g. dry weather, fog, rain, snowfall) Station is currently located, causing the transmit / receive and Interference range can be estimated even better.

<Communication requirement code <:

Definable, optional code that specifies which current or expected communication needs Regarding the type of service (stream, datagram)), scope (data volume) and characteristics (e.g. continuous, bursty traffic etc.) one station currently has.  

This makes it possible to use specified ones (assigned or variable) channel areas, ent speaking of the communication needs of a station, the existing capacity of the transmission medium in particular to use economically and the channel change rate too minimize.

<Validity time <:

Time until it is expected that for the knowledge-based method according to the invention for the medium specified parameters / codes (frequency, slot, Codeword identifier, transmission power, antenna and environment code) are valid.

Each station manages and updates a "direct neighbor" Table ", both based on those contained in the cluster table Information is based as well as current connectivity Stations within the transmission / reception range too ("Direct Neighbors ") and contains the following parameters:

Direct neighbor table

<Registration number <:

Sequence number of the entry.

<Station identifier << cluster identifier <:

Dynamic address of the direct neighbor of the respective Entry.

The current cluster identifier shows in which cluster the immediate neighbor is currently.

<Connection quality <:

This parameter gives the quality of the connection (English "Link Quality") to the direct neighbor.  

A connection quality of 0 indicates that the direct Neighbor only included in the direct neighbor table because, according to the cluster table, it is located within the same or an adjacent cluster; others if the value of this parameter is given by statistical evaluation of the emp catch signals.

Only in the event that the cluster Table currently saved parameters no station in located in the same or an adjacent cluster Stations in clusters with ditance 2 (i.e. an adjacent Cluster without station is in between) as a direct neighbor barn with a connection quality of 0 in the direct Neighbor table are included.

Are identified by one with connection quality 0 station (even despite polling) no signals received catch the station from the direct neighbor Table is deleted and it is replaced by if necessary Exchange of information with other neighboring stations (based on their neighbor tables and routing table) a local routing table or an existing one added.

<Direction of movement <:

Physical or logical direction of movement of the station according to the cluster table.

This parameter is used for directional routing used within the cluster. (For further explanations see the corresponding explanations in the cluster Table).

<Pass interval <or <Pass-x interval <, <Pass-y interval <, <Pass-z interval <:

These optional parameters (according to the cluster Table), are used to create stable routes across Neighboring stations moving at a speed move as close to your own speed as possible choose. (For further explanations see the corresponding the explanations in the cluster table.)

<Priority <:

This parameter indicates the priority with which the direct neighbor as mediation (relay) for communication cation with a remote station is selected.

The priority can be based on the following (or another reasonable) gradation can be determined:

The priority depends on:

• - The distance between the cluster in which the (own) station is currently located and the cluster in which the direct neighbor is located, with the following priority levels:
  • 3: Distance 1 (adjacent cluster)
  • 2: Distance 0 (same cluster)
  • 1: distance <1 (non-adjacent cluster);
• - the connection quality (e.g. 4 ... 0);
• - the relative speed (determined from <pass (x, y, z) interval <and direction of movement) with the following priority:
  • 4: Neighbor with approximately the same speed and direction
  • 3.5: Approaching neighbor with low relative speed
  • 3: moving neighbor with low relative speed
  • 2.5: Approaching neighbor with a moderate relative speed
  • 2: moving neighbor with moderate relative speed
  • 1.5: Approaching neighbor with great relative speed
  • 1: moving neighbor with great relative speed;
• where the value of the priority is the sum of the Individual gradations result.

The priority will depend on the last one Update the entry and / or the current one Priority dynamically reduced. With a priority of The entry (i.e. the direct neighbor) is deleted zero.

<Update time <:

Time at which the entry was last updated.

For the routing decisions according to the fiction according to knowledge-based adaptive route selection, each station manages and updates a "cluster routing Table "based on the cluster table and the following Parameter contains:

Cluster routing table

<Registration number <:

Sequence number of the entry

<Cluster identifier <:

Identification of the (target) cluster via the in the respective entry Routing information is stored.

<Route status <:

This status code ("Available", "Not available", "Unsure") indicates whether there is currently a way (route) to the corresponding clusters, based on the in the knowledge-based cluster table stored parameters exists:

Route available

It is assumed that a route to the cluster is available is when a continuous path over adjacent Clusters in which there is at least one station (due to the cluster table) exists.

Route unsafe

A continuous path across adjacent clusters in which at least one station is currently (on due to the parameters stored in the cluster table) not available, but it could only (one or more times) a cluster without a station between two clusters with at least least one station can be determined, so that it is because of the specified tolerance values and the actual Sending / receiving ranges of the stations, which over the (der Cluster dimension)) as a rule, it is possible that a physical route to the target cluster exists.

Route not available

Because of the stored in the cluster table Parameters, contains every possible route to the target cluster adjacent clusters at least two adjacent clusters in where there is currently no station.

<Route time <:

Time at the (due to that in the cluster table stored parameters) it is expected that:

• - one marked as "not available" (or "insecure") Route becomes "available" or  
• - a route marked as "available" is "not available" (or "unsafe").

<Distance <:

Number of clusters along the route to the target cluster.

<Runtime <:

Average signal / packet transit time to the target cluster. This Parameter is taken from the runtime of the cluster update messages and / or other data packets from stations calculated in the corresponding cluster and in the selection of the optimal route.

<Neighbor cluster <:

Cluster identifier depending on the route (to the destination cluster) to the cluster of the own station directly adjacent clusters.

This parameter is used to find a suitable one "Nearest neighbor" (see below) from the direct neighbor Table to choose.

<Nearest neighbor <:

Nearest (direct neighbor) on the route to the target cluster.

A station that specifies on the basis of its <function code < neither a relay nor an intermediary function fulfill is not set as <nearest neighbor <.

<Minimum pass list <:

List of expected minimum transit times (time interval vallen) depending on the direction of movement that the Stations due to permanent or temporal speed  Limitations on passing through the cluster will need.

This allows according to claim 2 of the invention the sending of update messages by stations the clusters with speed limits (maximum speed to be observed) drive through the at the travel speed intended by a station be avoided.

The values of this list can be used when building the cluster Routing table of a digital traffic route map ent taken or by querying other stations or Databases are determined. The values are set to Off send corresponding update messages Stations that pass through the cluster, as well as required If necessary updated by special queries.

<Mean passport list <:

List of the current average and the Variance of intended cruising speeds and actual speeds of the stations in Dependence of the direction of movement for the concerned Cluster.

This can according to claim 3 of the invention by estimating the expected length of time for the actual driving through the cluster depending on the Direction of movement and the intended travel speed the stations sending out updates reports by the stations (despite deviations from their intended cruising speed) avoided will.

The values of this list can be used when building the cluster Routing table by querying other stations or Databases are determined. The values are at Send out corresponding update messages  through stations that pass through the cluster, as well updated by special queries if necessary.

<Update time <:

Time the entry was last updated. Exceeds the difference to the current time a fixed value, the entry will be deleted.

With multi-level hierarchical subdivision according to the claim 4 of the invention is the cluster routing table hierarchical structure.

That to update the described knowledge-based Method used according to the invention in tables following described and explained.

Knowledge tables updated Update of the knowledge-based cluster table

Every time a station updates a cluster message received from another station is checked whether there is already an entry in the cluster Table is located. If not, a new entry is created and those contained in the cluster update message Parameters (<station identifier <, <cluster identifier <, (<position <, <Speed <, <acceleration <, <weight speed <), <Direction of movement <, <Send time <, <Next (x, y, z) arrival <, <Pass (x, y, z) interval <, <validity cluster <, <relevance zone code <, <function code <, (<frequency identifier <, <slot ID <, <code word ID <, <transmission power <, <antenna Code <, <environment code <, <communication requirement code <, <Validity period <)) are saved, otherwise the Entry with the cluster update received in the updated (new) parameter values.  

The cluster table is checked very frequently. If the current time is equal to or greater than that for one Entry as <next (x, y, z) arrival <stored value the <cluster identifier <of the entry to that of the direction of motion accordingly set neighboring clusters, and the value of the parameter <Next (x, y, z) arrival <:

• - by the value of <pass (x, y, z) interval <of the entry increases if the <function code <is a behavior of the Indicates station according to claim 1 of the invention)
• - by the maximum of the value of <pass (x, y, z) interval < for this entry and that in the cluster routing table contained value of <minimum pass list <for the corresponding clusters and the corresponding direction of movement if the <function code <is a behavior of the Station according to claim 2 of the invention)
• - the expected time to pass the Clusters (which result from the value of <pass (x, y, z) interval < for this entry and that in the cluster routing table contained value of <mean list <for the ent speaking clusters and the corresponding direction of movement is determined), increased if the <function code <on Behavior of the station according to claim 3 of Invention indicates).

When the current <cluster identifier <reaches a value outside the value of the <validity cluster <or outside the relevance zone indicated by the <relevance zone code < the entry is deleted from the cluster table.

If the difference between the current time and the saved (last) <update time <one set value exceeds the entry from the cluster Table deleted.

Update of the direct neighbor table

The direct neighbor table is then updated each time (or checked) if:

• - there was a change in the cluster table; the the corresponding entry will only be updated if the direct neighbor with a connection quality of zero is expelled or a new direct neighbor to the Direct neighbor table is included.
• - a message from a direct neighboring station (data packet) received: <connection quality <and <priority < are determined, <cluster identifier <, <update time <, as well as any transmitted parameters are updated to the current len values set.
• - The priority of an entry must be reduced.

Update of the cluster routing table

The cluster routing table is checked and if necessary updated when:

• - there was a change in the cluster table:
  <Route status <, <Route time <, <Distance <, <Running time <and <Neighbor cluster <are taking into account the shortest distance (or the <connection quality <to direct neighbors at the same distance) or the shortest running time determined and set to the current values.
• - there was a change in the direct neighbor table:
  The value of the parameter <nearest neighbor <is left unchanged if no other neighboring station with a higher priority is stored in the direct neighbor table; otherwise the station with the highest priority is selected as the <nearest neighbor <(or one station at random with several stations with the same priority).
• - Receive an update message containing new values for the <minimum pass list <(according to claim 2 of the invention) and / or new values regarding the <mean pass list <(according to claim 3 of the invention) from a station has been:
  The corresponding entries in the cluster routing table are updated.

Each time an entry is updated, the <update time <set to the current time.

Knowledge-based adaptive route selection - routing decision

To establish a connection to a destination station place or a data packet there via a neighboring station forward

• - according to the address (<station identifier << cluster- ID <) of the target station currently in the cluster Table stored for this station <cluster identifier < detected;
• - if this <cluster identifier <with that in the address of the Target station contained <cluster identifier <does not match true, the target station is given a new address (dynamic, knowledge-based addressing (!)) according to the in the Cluster table stored value assigned and in Data packets used as new destination address;
• - then for the corresponding <cluster identifier < Address of the <nearest neighbor <in the cluster routing table determined and the data packet to this neighboring station Posted.

If there is currently no route due to the cluster routing table the cluster in which the target station is currently located, is available due to the knowledge-based Cluster table stored information are calculated when the target station will eventually reach a cluster  then a route exists so that data packets only then and also be sent at the right time, even if that It is very likely that reaching the destination station itself, as a further advantage of the knowledge-based process of Invention, a significant reduction in network load results.

Through the method of knowledge-based, dynamic addressing It is also possible to use stations that are not inventive appropriate procedures for adaptive route selection but others Use routing procedures in the communication network too integrate.

Cluster update messages

According to the method for knowledge according to the invention based adaptive route selection (routing) and knowledge based Control of access to the transmission medium are new additional stations or stations the deviations of the in parameter values stored in their knowledge-based tables actually measured or recalculated values below Determine their current position, "Cluster update messages" sent out to the contain current parameter values.

If measured or calculated values for certain parameters exceed the specified maximum values, the cluster Update messages the corresponding maximum values used.

The cluster update messages are of variable length, that of the specified accuracy and range of values Parameters, the variant (according to claims 1, 2 or 3 of the Invention) depends and contain the following fields, where the fields in square brackets are optional can be sent:

<Cluster update message of (main) variant 1 <:: =
<Station identifier << cluster identifier <[<position <
<Speed << acceleration << weight speed <]
<Direction of movement << Send time << Next (x, y, z) arrival <
<Pass (x, y, z) interval << validity cluster <
<Relevance zone code << function code <
[[<Frequency identifier <] [<slot identifier <] [<code word identifier <]
<Transmission power <[<Antenna code <] [<Environment code <]
[<Communication requirement code <] <Validity period <]

<Cluster update message of (main) variant 2 <:: =
<Station identifier << cluster identifier <[<position <]
<Speed << acceleration << weight speed <]
<Direction of movement << Send time << Next (x, y, z) arrival <
<Pass (x, y, z) interval << validity cluster <
<Relevance zone code << function code <
[[<Frequency identifier <] [<slot identifier <] [<code word identifier <]
<Transmission power <[<Antenna code <] [<Environment code <]
[<Communication requirement code <] <Validity period <]
<Minimum pass list <

<Cluster update message of (main) variant 3 <:: =
<Station identifier << cluster identifier <[<position <]
<Speed << acceleration << weight speed <]
<Direction of movement << Send time << Next (x, y, z) arrival <
<Pass (x, y, z) interval << validity cluster <
<Relevance zone code << function code <
[[<Frequency identifier <] [<slot identifier <] [<code word identifier <]
<Transmission power <[<Antenna code <] [<Environment code <]
[<Communication requirement code <] <Validity period <]
[<Minimum pass list <] <mean value list <

Significant advantages of the method according to the invention for knowledge-based adaptive route selection and knowledge-based Control of access to the transmission medium opposite  known methods are that described Update messages (the important information about the Fulfillment of the functions of the network and switching level included), as a rule (i.e. in the case of something that is not too changeable Mobility behavior of the stations) only relatively rarely must be sent out and disseminated, and that nevertheless good and extensive information for everyone Stations about the current and even a prediction about the future network topology, also with communication networks with a large number of stations with high mobility, both for the implementation of efficient adaptive routing Procedure as well as for a collision-free as possible Access control and economic use of the Transmission medium is crucial.

Claims (8)

1. Knowledge-based method for adaptive route selection (routing), dynamic addressing and control of access to the transmission medium (media access control) in a communication network with the following features:

• - The communication network consists of mobile stations or of mobile stations and fixed stations, the information being transmitted between the stations via suitable transmission media;
• - Communication with mobile stations takes place either via fixed stations, in the transmission / reception area of which the mobile stations are moving, or via other mobile stations which act as a relay (relay) if the communicating stations are not themselves the source or sink of the transmitted information are;
• - The mobile stations can determine or estimate their own current position, speed and direction of movement;
• the stations can determine or estimate the current time with an accuracy of at least one second;

characterized by the following features:

• - For route selection, addressing, medium access and other functions, the communication network system-wide is based on a division into fields known in the stations (hereinafter referred to as "cluster"), this being done according to geographical or logical aspects, and in the dimensioning of the clusters the transmission / reception radii of the stations are taken into account;
• - Each station continuously manages and updates the information required for route selection, addressing and medium access via all stations of the communication network or only part of the same in a station-specific determinable surrounding area of a station (hereinafter referred to as "relevance zone") using digital tables or The like, which are called "knowledge-based tables" in the following, so that the cluster in which a mobile station is currently located can be determined virtually without delay and, moreover, a prediction of the achievement of a particular cluster at a future point in time is possible;
• - For the quasi-synchronous update of the knowledge-based tables managed decentrally by the stations, each newly added station exchanges the information required for the construction or expansion of these tables with neighboring stations, which if necessary are further disseminated by the participating stations within the relevance zone ;
• - Each station constantly compares the information it sends out, which is used to update the knowledge-based tables managed by the stations, with the actual values and, in the event of a deviation beyond the defined tolerance range, sends the current values to all neighboring stations, if any required to be further disseminated within the relevance zone;
• - The stations are addressed dynamically and, in addition to the station identifier, contains the identifier of the cluster in which the station is currently located;
• - For route selection (English "routing"), the knowledge-based tables, which are managed decentrally and are updated virtually synchronously, are evaluated in order to determine the cheapest route (route) to a station in a particular cluster;
• - For access to the transmission medium, the knowledge-based tables, which are managed decentrally and are updated virtually synchronously by the stations, are evaluated in order to make optimum use of the transmission medium and to avoid access conflicts.

2. Knowledge-based method according to claim 1 with the additional ability of all or individual stations in speed clusters in certain clusters knowing or being able to query, characterized in that the stations in updating the regarding this Information expanded knowledge-based tables that too expected maximum speed of a station when passing through of the cluster, thereby sending out Messages with current values (because of the Speed limit) by the corresponding station is largely avoided. 3. Knowledge-based method according to claim 2 with the additional ability of all or individual stations Mean and the standard deviation of the intended Cruising speeds of the stations in a cluster (under Taking into account the direction of movement) and the mean and the standard deviation of the actual speeds can determine or query, characterized in that the Stations when updating regarding this Information expanded knowledge-based tables that too expected speed of a station when passing through the Consider clusters, thereby sending out messages largely with current values by the corresponding station is avoided.   4. Knowledge-based method according to claim 1, 2 or 3, characterized in that the communication network system-wide a multi-level hierarchical division into fields (with appropriate clustering on different Hierarchy levels) is used. 5. Knowledge-based method according to claim 1, 2, 3 or 4, characterized in that the communication network system-wide a three-dimensional subdivision into spatial areas appropriate clustering is used. 6. Knowledge-based method according to claim 1, 2, 3, 4 or 5, characterized in that the stations have the property either independently or at the request of others Stations with control function their functionality as Switching station (with full routing capability), Relay station (retransmission of received messages) as Station with control function and / or information function (under Consideration of necessary, defined requirements) or as a station without one of the functions mentioned and can also change dynamically. 7. Knowledge-based method according to claim 1, 2, 3, 4, 5 or 6, characterized in that not all stations are perfect manage the knowledge-based tables decentrally and update, but managing and updating the knowledge-based tables by next to or exclusively existing, equal or hierarchical order subject control stations with control and / or information function takes place to which the other stations with less Functionality and any restricted properties Update messages to build the knowledge-based Send tables or get the information you need.