JP2007221568A - Transfer delay control method and radio terminal - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To guarantee a delay of end-to-end data transfer in a network for performing a communication mutually after a plurality of radio terminals are performed self-organization. <P>SOLUTION: A plurality of routs are to be searched between a transmission source terminal and a designated terminal relating to data to be transmitted, then one of a plurality of the searched routes is to be transmitted after being selected. Transmission priority is to be raised relating to the data when transfer delay cannot satisfy a request while transfer route is to be switched when the request does not be further satisfied, and route search is to be retried when the request does not be still further satisfied. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to end-to-end data transfer within a wireless network. In particular, the present invention relates to a technique for guaranteeing a delay in data transmission of real-time traffic.

  The present invention is particularly suitable for use in ad hoc networks. An “ad hoc network” is a self-distributed wireless network in which wireless terminals such as personal computers, mobile phones, and personal digital assistants (PDAs) are self-organized. There is a similar term “multi-hop wireless network”, but this includes a case where a fixed facility such as a base station is interposed. In the present invention, an “ad hoc network” is limited to a pure distributed network in which terminals communicate with each other without any fixed facilities.

  As technologies for constructing an ad hoc network, CSMA / CA (Carrier Sense Multiple Access with Collision Avoidance) protocol for enabling multiple access and DSR (Dynamic Source Routing) algorithm for multi-hop communication (non-patent literature) 1) and AODV (Adhoc On-Demand Distance Vector) routing algorithm (see Non-Patent Document 2), many wireless technologies, protocols and algorithms have been studied.

However, until now, little research has been conducted on techniques for ensuring quality of service (QoS) between end-to-end within an ad hoc network. In fact, securing end-to-end QoS is incompatible with the dynamic topology and mobility inherent in ad hoc networks, and is a conventional method used in wired networks and cellular networks, such as destinations on the network. Resource reservation that reserves the bandwidth up to this point and secures communication quality cannot be used in an ad hoc network as it is.
Internet Draft, “The dynamic source routing protocol for mobile ad hoc networks (DSR).”, Http://www.ietf.org/internet-drafts/draft-ietf-manet-dsr-10.txt RFC3561, “Ad hoc on demand distance vector (AODV) routing.”, Http://www.ietf.org/rfc/rfc3561.txt IEEE802.11e “Part 11: Wireless Medium Access Control (MAC) 6 and Physical Layer (PHY) specifications: Medium Access Control (MAC) Enhancements for Quality of Service (QoS)”

  In order to ensure end-to-end QoS in ad hoc networks, it is essential to dynamically modify resources so that a particular performance metric is maintained below or above a predetermined threshold. Become.

  As a method for dynamically correcting resources, it is conceivable to modify mapping of various radio resources such as transmission speed, transmission range, transmission probability, and the like so as to suit the radio environment. However, each of these parameters is related to a specific access technology or physical technology, and it can be modified only in a specific usage pattern. For example, the modification of the transmission range is suitable for a network in which a terminal can change its transmission power in various transmission ranges, but a network with a fixed transmission power such as an ad hoc network standardized in the IEEE 802.11 series. That is, it cannot be adopted in a network having a fixed transmission power in which the transmission range and transmission power are fixed in all terminals and transmissions. In order to design a robust protocol, it is necessary to develop common parameters for adaptively modifying the wireless environment.

  In an ad hoc network, even if QoS is secured between links or points, it is not always possible to secure QoS between end and end. For example, a QoS (Media Access Control) protocol such as EDCF defined in IEEE802.11e (see Non-Patent Document 3) can guarantee link quality. However, the end-to-end path quality configured by combining links is not guaranteed by the QoS MAC protocol of each terminal. To ensure end-to-end QoS with the QoS MAC protocol, all terminals in the path need to cooperate with each other. Moreover, it is very important how to allocate the individual QoS parameters of the link in order to satisfy all end-to-end QoS requirements within the dynamic network. The QoS MAC protocol requires QoS parameters such as a priori defined delay or speed. For example, how to divide end-to-end delays into link delay requirements. On the other hand, since the optimum division of delay varies depending on the radio environment, it is necessary to dynamically modify the division, which is very difficult.

  Routing algorithms have also been developed to find routes or paths that satisfy end-to-end QoS requirements, but these algorithms generally rely on complex route maintenance algorithms and network resource management protocols. For this reason, how much adaptive correction is possible is limited to these algorithms.

  The present invention solves such problems and provides a transfer delay control method capable of guaranteeing a delay in data transfer between end and end without depending on various algorithms or protocols, and a wireless terminal therefor. Objective.

  The transfer delay control method of the present invention is a method for guaranteeing transfer delay between end-to-end on a network in which a plurality of wireless terminals are self-organized and communicate with each other. A first step of searching for a plurality of routes with a destination terminal; a second step of selecting one of the searched plurality of routes and transmitting the data to be transmitted from the source terminal to the destination terminal; And a third step of increasing the transmission priority for the data to be transmitted when the transfer delay from the transmission source terminal to the destination terminal does not satisfy the request, and increasing the transmission priority. However, when the request is not satisfied, another route out of the plurality of routes searched in the first step for a transfer route from the transmission source terminal to the destination terminal Switching first includes fourth and steps, when the even switching the transfer route by the fourth step the request is not satisfied, characterized in that to start from the first step.

  In the third step, when the transfer delay for the data to be transmitted is smaller than necessary, it is desirable to lower the transmission priority for the data.

  Specifically, a traffic level is set for each data transferred on the network according to the real-time property required for the data, and in the third step, the traffic level of the data to be transmitted is set. It is determined whether or not the transfer delay is satisfied. Each traffic level is assigned a maximum amount of delay that the level of data can tolerate and a minimum amount of delay that is not required until a smaller delay. One or more transmission priorities are individually assigned to each traffic level.

  Before transmitting the data to be transmitted in the second step, it is desirable to transmit pseudo data for confirming whether the selected route can actually satisfy the transfer delay request.

  According to the present invention, there is further provided a wireless terminal used in the above network. That is, the wireless terminal includes a means for searching for a route with a destination terminal on the network and a means for transmitting data to the searched route. The means for searching includes a plurality of routes. Means for searching and storing in a route table, wherein the means for transmitting increases the transmission priority for the data when notified from the destination terminal that the transfer delay request for the transmitted data is not satisfied, Alternatively, it is characterized by including means for switching the transfer route to another route stored in the route table when it cannot be raised further.

  The means for transmitting preferably includes means for lowering the transmission priority for the data when notified from the destination terminal that the transfer delay of the transmitted data is less than necessary.

  Means for receiving data from the source terminal and measuring the transfer delay; and means for notifying the source terminal of the measured transfer delay if it does not satisfy the range required for the data; It is desirable to provide further.

  In the present invention, transmission priority is given to data to be transmitted, and when the required transfer delay cannot be satisfied, the transmission priority is increased. If you are still not satisfied, switch the route. The transmission priority is not a special QoS scheme, but can be adapted to various protocols and algorithms to set the priority of traffic in the MAC or physical layer. As a result, it is possible to satisfy the required transfer delay between end and end within the range of the system capacity without needing to perform unnecessary route search.

  It should be noted that “transmission priority” in the present invention is not a parameter directly related to the transfer delay required for data to be transmitted, but a parameter to be corrected when the transfer delay is not satisfied. It is. Therefore, as long as the transfer delay is satisfied, data can be transferred with a low priority, and resources can be efficiently shared among users by using the resources for other data transfers.

  On the other hand, a traffic level is set for each data as a parameter directly related to the transfer delay required for the data to be transmitted. Each traffic level is assigned one or more transmission priorities. At this time, a maximum transmission priority different from other traffic levels is assigned to each traffic level according to a transfer delay required for the traffic level. By assigning a high traffic level to data such as voice and video that require real-time performance, it is possible to obtain a higher transmission priority than other data even during congestion, and to guarantee transfer delay as much as possible. Can do.

  Further, when the transfer delay is unnecessarily small, the transmission priority for the data is lowered, so that resources can be efficiently shared between users, and the performance of the entire network can be improved.

  FIG. 1 is a diagram showing a configuration example of a wireless network implementing the present invention, and FIG. 2 is a diagram simply showing a procedure of a transfer delay control method according to the present invention.

  This wireless network forms an ad hoc network, and wireless terminals 11 to 16 such as personal computers, mobile phones, and portable information terminals are self-organized and communicate with each other. FIG. 1 shows an example in which data from the wireless terminal 11 to the wireless terminal 15 is transferred via the wireless terminal 13.

  Here, a method for guaranteeing an end-to-end transfer delay will be described with reference to FIG. 2 as an example in which data is transferred from the wireless terminal 21 to the wireless terminal 26. A plurality of routes are searched for between data to be transmitted between the source wireless terminal 21 and the destination wireless terminal 26. Here, it is assumed that a route passing through the wireless terminals 23 and 24 and a route passing through the wireless terminal 25 are found. The transmission source wireless terminal 21 selects one of the searched routes, in this case, the route passing through the wireless terminals 23 and 24, and transmits the data to the wireless terminal 26 in a packet. The wireless terminal 26 obtains the average delay of the received data, and if the value does not satisfy the request, notifies the wireless terminal 21 of it. When receiving the notification from the wireless terminal 26, the wireless terminal 21 increases the transmission priority for the data. If the request is not satisfied even if the transmission priority is raised, the transfer route from the wireless terminal 21 to the wireless terminal 26 is switched to another route, in this example, the route via the wireless terminal 25. If the delay request is still not satisfied, the route is searched again.

  Detailed examples will be described below.

  FIG. 3 is a flowchart of a method for ensuring end-to-end delay using a priority model according to the present invention. Although a simple routing algorithm is illustrated here, the present invention is not limited to the routing algorithm, and other complicated or efficient routing algorithms may be used.

  When the transmission source terminal wants to transmit data, it first transmits a request message PREQ for inquiring about the route (S11). This PREQ has a destination address and a transmission time, and is transferred by every possible route. As soon as the destination terminal receives the PREQ, it sends back a confirmation message RREP (S21), and is transferred by the reverse route to the PREQ. On the other hand, RREP uses the time information attached to the received PREQ in order to estimate the route delay. When the transmission source receives the RREP, the route information attached to the RREP is obtained and the route is stored in the transmission source route table (S12). However, the transmission source does not immediately use this route to transmit data, but waits for a certain period of time until it receives an RREQ with other route information for the backup route. At the same time, the transmission source compares the time information attached to the RREP with the reception time of the RREP, and calculates the delay of the route. When the waiting time has passed and the route table includes at least one route, the source transmits data using an appropriate route in consideration of delay constraints.

  According to our analysis and simulation, it has been observed that RREQ / RREP packets are subject to different delays than data packets in various networks and access protocols. Detailed differences are currently unknown. Therefore, the source terminal transmits pseudo data with the maximum priority allowed for the data in order to check the delay of the data packet (S13). The reason for using the maximum priority is to confirm whether or not the delay request can be satisfied by the route. The destination terminal that has received the pseudo data returns a confirmation message ACK (S22). When the reception of the pseudo data is confirmed by the returned ACK, the transmission source terminal calculates whether the delay of the route is satisfactory for the data (S14). Then, the most appropriate route is selected from the routes satisfying the data delay request (S15), and the data is transmitted with the lowest priority (S16). If there is no route that satisfies the request, the process is repeated from finding a new route.

  Network conditions change dynamically and traffic needs to adapt itself to the network environment. Therefore, the destination terminal continues to record the traffic delay and calculates the average traffic delay (S23). When the average delay is larger than the delay request or smaller than the minimum threshold (S24), the destination terminal transmits a Mark ACK to the transmission source terminal (S25). After receiving the Mark ACK, the source terminal modifies the data transmission priority (S17). If the priority correction cannot satisfy the QoS request, the source terminal selects another route from the route table (S19, S110), and transmits data (S16). Details of the priority correction will be described later.

  FIG. 4 shows the relationship between traffic level and transmission priority. In the present invention, the traffic level is set for each data transferred on the network according to the real-time property required for the data. Each traffic level is individually assigned one or more transmission priorities. Assuming that there are M traffic levels and N priority levels, in the prior art, one priority corresponds to one traffic level and N = M. In the present invention, N> M.

  Each traffic level has a maximum amount of delay that the data at that level can accept (this is called the “maximum delay constraint”) and a minimum threshold amount of delay that is not required until a smaller delay (this is "Minimum delay constraint"). FIG. 5 shows the relationship between the traffic level and the delay constraint. This will be described in detail later.

  In FIG. 4, the minimum transmission priority of each traffic level is the same value “1”. This means that data is transmitted with the same minimum transmission priority “1” for each traffic level. However, the present invention is not limited to the same minimum transmission priority for each traffic level, and a different minimum transmission priority may be assigned for each traffic level.

  The transmission priority of each data can take any value between the minimum transmission priority and the maximum transmission priority as long as the delay constraint is satisfied. According to the analysis by the present inventors, the transmission priority is kept low when the traffic load is light, and the transmission priority is close to the maximum transmission priority when the traffic load is high. Therefore, by setting the minimum transmission priority to the same “1” for all traffic levels, when the traffic is light and the resources are sufficient, the resources can be occupied by the low priority traffic.

  As shown in FIG. 5, one maximum delay constraint and one minimum delay constraint are set for each traffic level. The maximum delay constraint is a parameter conventionally used for a difference in traffic. For example, the maximum delay constraint for voice or video traffic is much smaller for other data traffic. In contrast, the minimum delay constraint is not for traffic differences but for optimal allocation of resources. When the average traffic delay is less than the minimum delay constraint, it means that the traffic is consuming more resources than the request and its priority is a little higher or too high. Therefore, the transmission priority of data for the traffic is reduced, and at the same time, an acceptable delay is maintained. Indirectly released resources can be used by other traffic to achieve optimal resource allocation. In FIG. 5, it is assumed that the minimum delay constraint is smaller than the maximum delay constraint by the same value at each traffic level, and the difference is small. The present invention is not limited to such a case, and optimal allocation of the maximum delay constraint and the minimum delay constraint should be set as appropriate.

  FIG. 6 is a flowchart showing details of correction of transmission priority. When the transmission source receives the Mark ACK, the transmission source modifies the transmission priority of the data in accordance with the information conveyed by the Mark ACK. First, when the average delay calculated by the destination terminal and transmitted in the Mark ACK packet is larger than the maximum delay constraint (S31), if the transmission priority is lower than the maximum priority at the traffic level of the data (S32), the transmission source The terminal increases the transmission priority for the data (S33). Otherwise, another route is searched from the route table (S37), and if a spare route exists, it is selected (S38). When the average delay is smaller than the minimum delay constraint (S34), if the transmission priority of the data is greater than the minimum priority (S35), the transmission source terminal lowers the transmission priority of the data (S36). By lowering the transmission priority, it is possible to release the excessive resource occupancy caused by high priority traffic and balance the traffic load.

  In order to transmit data according to the transmission priority in each wireless terminal, it is possible to use a conventional technique such as providing a buffer for each priority and sequentially transmitting data stored in a buffer with a high priority. . However, in the past, when a buffer having a different priority was provided corresponding to the maximum delay constraint of each data and the correspondence between the type of data and the buffer was fixed, this technique is used in the present invention. Does not have a direct correspondence between the data type (traffic level) and the buffer, and the data is transferred to the corresponding priority buffer according to the transmission priority set adaptively for each data. Try to distribute.

The figure which shows the structural example of the wireless network which implements this invention. The figure which shows simply the procedure of the transfer delay control method by this invention. Flowchart of a method for ensuring end-to-end delay. The figure which shows the relationship between a traffic level and transmission priority. The figure which shows the relationship between a traffic level and a delay constraint. The flowchart which shows the detail of correction of transmission priority.

Explanation of symbols

  11-16, 21-26 Wireless terminal

Claims (9)

In a transfer delay control method for guaranteeing end-to-end transfer delay on a network in which a plurality of wireless terminals are self-organized and communicate with each other,
A first step of searching for a plurality of routes between a source terminal and a destination terminal for data to be transmitted;
A second step of selecting one of a plurality of searched routes and transmitting the data to be transmitted from the transmission source terminal to the destination terminal;
A third step of increasing the transmission priority for the data to be transmitted when the transfer delay from the transmission source terminal to the destination terminal fails to satisfy the request;
Fourth step of switching the transfer route from the source terminal to the destination terminal to another route among the plurality of routes searched in the first step when the request is not satisfied even if the transmission priority is increased. Including and
A transfer delay control method, comprising: starting from the first step when the request is not satisfied even if the transfer route is switched in the fourth step.   The transfer delay control method according to claim 1, wherein, in the third step, when a transfer delay is smaller than necessary for the data to be transmitted, the transmission priority for the data is lowered. A traffic level is set for each data transferred on the network according to a real-time property required for the data,
The transfer delay control method according to claim 1, wherein in the third step, it is determined whether a transfer delay is satisfied with respect to a traffic level of the data to be transmitted.   4. The transfer delay control method according to claim 3, wherein each traffic level is assigned a maximum delay amount that can be accepted by the data of the level and a minimum delay amount that is not required until a smaller delay.   4. The transfer delay control method according to claim 3, wherein one or a plurality of transmission priorities are individually assigned to each traffic level.   2. The transfer delay control method according to claim 1, wherein before transmitting the data to be transmitted in the second step, pseudo data for confirming whether the selected route can actually satisfy the transfer delay request is transmitted. Means for searching for a route with a destination terminal on the network;
In a wireless terminal equipped with means for transmitting data to the searched route,
The means for searching includes means for searching a plurality of routes and storing them in a route table;
When the transmission means is notified from the destination terminal that the transfer delay request for the transmitted data is not satisfied, the transmission means increases the transmission priority for the data, or if the transmission priority cannot be further increased, A wireless terminal comprising means for switching a transfer route to another route stored in a route table.   8. The wireless terminal according to claim 7, wherein the means for transmitting includes means for lowering the transmission priority for the data when notified from the destination terminal that the transfer delay of the transmitted data is unnecessarily small. Means for receiving data from the source terminal and measuring its transfer delay;
The wireless terminal according to claim 7 or 8, further comprising: means for notifying the transmission source terminal of the measured transfer delay when the data does not satisfy a range required for the data.

Images