KR101424228B1 - A Path Selection Method for distributed scheduling in Mobile Multi-hop Relay Networks - Google Patents

Abstract

The present invention relates to a path selection method for distributed scheduling of a mobile multi-hop relay network, and more particularly, to a path selection method for distributed scheduling of a mobile multi-hop relay network for a broadband wireless access including a plurality of relays, (RS) broadcasts a medium access control message having a parameter value for path selection; Scanning a medium access control message broadcasted by a new RS (New RS) entering the MMR network and generating a path metrics table based on parameter values for path selection among the received MAC access control messages; Calculating a path cost for each of a path between a new relay (New RS) and a base station (BS) in the MMR network based on information in the path metrics table; A path selecting unit that selects a relay unit of the path having the most optimal cost among the calculated path costs as a parent relay unit of a new RS and selects a BS connected to the parent relay unit as a serving BS of a new RS And a selection step. This provides the effect of allowing each repeater to select the path between the repeater and the base station having the highest achievable data throughput by a distributed scheduling scheme.

MMR, distribution, scheduling, path, selection, repeater

Description

[Background Art] [0002] A path selection method for distributed scheduling of a mobile multi-hop relay network {

The present invention relates generally to a mobile multi-hop relay (MMR) network proposed by IEEE (Institute of Electrical and Electronics Engineers) 802.16j, and more particularly to a multi- A method for selecting a path for distributed scheduling in a mobile multi-hop relay network, which allows each repeater to select a path between a repeater and a base station having the highest data throughput rate that can be reached in a hop-based MMR network by a distributed scheduling scheme will be.

Recently, the IEEE 802.16 Relay Task Group is developing a new standard for mobile multi-hop relay (MMR) for broadband wireless access at high speed and low cost. The IEEE 802.16j MMR system is intended to extend the service coverage of a base station and improve data throughput and reliability of a communication link in a broadband wireless access communication network such as a conventional Wibro / Wimax network. Also, the power of the terminal can be saved while the terminal is connected to the repeater located closer to the base station.

According to IEEE 802.16j, a repeater is defined in two modes according to the purpose of introduction. One is a transparent mode. Since the terminal is located in the base station area, the base station can directly receive control information such as broadcast UCD / DCD or DL-MAP / UL-MAP due to its low signal quality . In this case, the repeater delivers only the data channel to the mobile station and directly transmits the control information to the mobile station. The other is called a non-transparent mode, which is for a terminal that is located outside the base station area and can not receive service. Both the control channel and the data channel are relayed in the repeater, thereby enabling communication between the terminal and the base station. In this case, the repeater performs a process of initial network entry, ranging, and the like in place of the terminal.

When the new RS enters the existing communication network or when the mobile relay station moves into the network, the incoming repeater must perform the registration procedure with the existing repeater or the base station. In the 802.16e standard, the connection between the base station and each terminal is performed separately by the CID. On the other hand, in the MMR network, since one or more relays may exist between the BS and the MS, the routing of the RSs is required.

Repeater path setup is required to find and select the best path to the base station considering limited radio resources. Two known routing schemes for MMR networks are centralized routing and distributed routing. In 'distributed routing', each repeater independently provides information for path selection. In 'centralized routing', only the base station knows the topology of the MMR network and stores the route information. In this case, the base station can use source routing, and each repeater only performs the role of forwarding along the given path from the received data.

As the number of repeaters increases in the MMR network, it becomes difficult to manage the entire network by the 'centralized routing' method. Also, as the number of 'hops' increases, the communication delay between the terminal and the base station increases, and additional radio resources are needed when relaying the channel. Therefore, a distributed path selection technique based on 'distributed routing' is required for reducing communication delay and efficient use of radio resources in route selection.

Recently, several studies have been conducted on the MMR of 802.16. For example, the paper Ohyun Jo and Dong-Ho Cho "Traffic adaptive uplink scheduling scheme for relay station in IEEE 802.16 based multi-hop system", The IEEE 66 ^th Vehicular Technology Conference , Sept. 30 2007-Oct. 3 In 2007, we propose an efficient uplink scheduling method of repeater to improve bandwidth management and delay performance. However, in the paper by Ohyun Jo and Dong-Ho Cho, only the scheduling of the base station and the repeater is considered, and the problem of how each repeater selects the path in the MMR network composed of a plurality of repeaters is not discussed .

In the meantime, it is known, for example, that research on the path selection of nodes based on the existing 802.16 mesh is described in, for example, Hung-yu Wei, Samrat Ganguly, Rauf Izmailov, and Zygmunt Haas, "Interference-Aware IEEE 802.16 WiMax Mesh Networks" ^{the 61 st IEEE Vehicular Technology Conference,} May 2005 , and papers Yang Cao, Zhimin Liu, and Yi Yang, "A Centralized Scheduling Algorithm based on Multi-path Routing in WiMAX Mesh Network", Wireless Communications, Networking and Mobile Computing, Sept. 2006, and Nahle, S., Iannone, L., Donnet, B., and Malouch, N. "On the Construction of a WiMAX Mesh Tree", IEEE Communications Letters , December 2007, and the like, but they are also based on a centralized scheduling scheme. Therefore, in an MMR network composed of a plurality of repeaters, It is not mentioned.

Therefore, there is an urgent need for a new technique that allows each repeater to select an optimal route by a distributed scheduling scheme instead of a centralized routing scheme in a multi-hop based MMR network including a plurality of repeaters.

The present invention proposes an optimal path selection method for each repeater in a conventional IEEE 802.16 based mobile multi-hop relay (MMR) network. It minimizes communication delay in MMR network, A method of finding a path of a repeater capable of maximizing the number of repeaters, that is, a method in which each repeater in a MMR network including a plurality of repeaters calculates a route based on a combination of an available bandwidth, a signal-to- noise ratio (SNR), and a hop- It is an object of the present invention to provide a path selection method for distributed scheduling of a new mobile multi-hop relay network, which has a low path cost or selects an optimal path with the highest data throughput.

The above object is achieved by a path selection method for distributed scheduling of a mobile multi-hop relay network provided by the present invention.

A path selection method for distributed scheduling of a mobile multi-hop relay network provided according to an aspect of the present invention includes a mobile multi-hop relay for a broadband wireless access including a plurality of relays (RSs) A method of selecting a path for distributed scheduling of an MMR network, the method comprising the steps of: each repeater (RS) broadcasting a medium access control message having a parameter value for path selection; A new RS (New RS) entering the MMR network scans the broadcast medium access control message and generates a path metrics table based on the parameter values for the path selection among the received media access control messages; Calculating a path cost for each of the paths between the new RS (New RS) and the base station (BS) in the MMR network based on information in the path metrics table; Selects a repeater of the route having the most optimal cost among the calculated path costs as a parent relay of the new RS and transmits a BS connected to the parent relay to a serving BS of the new RS And a path selection step for selecting the path.

In a preferred embodiment, the mobile multi-hop relay (MMR) network for the broadband wireless access is based on IEEE 802.16j, the broadcast medium access message is a UCD (Uplink Channel Descriptor) The parameter values for path selection are coded in a TLV (Type-Length-Value) format.

In a preferred embodiment, in calculating the path cost for each path, the path cost (P) for each path is calculated by dividing the data throughput (L) of each path by the hop-number (H) Value, i.e., P = L / H.

Also, in a preferred embodiment, after the path selection step, the new RS periodically broadcasts a path parameter value to a child RS.

The present invention provides a method of minimizing a communication delay and maximizing throughput in a mobile multi-hop relay (MMR) network, that is, a method of finding a path of a repeater, Each repeater in the MMR network has the lowest path cost based on a combination of available bandwidth, SNR and hop-count, or selects an optimal path with the highest data throughput And so on.

Hereinafter, a detailed example of a path selection method for distributed scheduling of a mobile multi-hop relay network according to the present invention will be described in detail with reference to the accompanying drawings.

FIG. 1 is a schematic diagram illustrating a medium access message configuration used in a path selection method for distributed scheduling in a mobile multi-hop relay (MMR) network according to an embodiment of the present invention. FIG. 3 is a schematic diagram illustrating an MMR network in which a path selection method for distributed scheduling according to an embodiment of the present invention is used. FIG. 3 is a schematic diagram illustrating an example of an MMR network in which a path selection method for distributed scheduling according to an exemplary embodiment of the present invention is used. FIG. 4 is a schematic diagram illustrating a path cost table that may be generated based on an example Done MMR network in FIG. 2 where a path selection method for distributed scheduling according to an embodiment of the present invention is used , FIG. 5 is a schematic diagram illustrating a path cost table that can be generated in the path selection method for distributed scheduling according to an embodiment of the present invention, Figure 6 is a graph of the path selection method for distributed scheduling in accordance with one embodiment of the present invention, illustrating a data throughput in each path according to the position of the new repeater.

According to the IEEE 802.16e-2005 standard, an uplink channel descriptor (UCD) message is a MAC management message that periodically broadcasts at a base station to provide a burst profile. Parameter values of a MAC management message including a UCD message used in IEEE 802.16e are coded in a type-length-value (TLV) format for flexibility.

The traffic delay through a multi-hop relay depends on available bandwidth, SNR (signal-to-noise ratio) and hop-count. A new parameter value 10 for path selection can be added to the UCD message 1 while retaining the existing TLV encoded parameter values according to the present invention.

FIG. 1 shows that the parameters necessary for path selection are encoded in the TLV format and added to the existing UCD message. In multi-hop scenarios, the path cost between the repeater and the serving base station is a much more important consideration than the path cost between the repeater and the parent relay. Therefore, all repeaters and links located on the path to the serving base station as well as the link state with the parent repeater must be considered.

If the repeater examines the initialization procedure for initial entry into the base station, the repeater entering the network tries to acquire a strong synchronization with the existing network. If this fails, it continuously scans the available channels until a useful network is found. Once synchronization is successful in the PHY, the MAC attempts to acquire channel information and network parameters. The new repeater creates a table for neighboring repeaters, receives the UCD messages broadcast by neighboring repeaters, and stores the repeater path parameter values obtained therefrom. If the same UCD message is obtained twice from the neighboring repeater, the neighboring repeater list formation is terminated. This means that the neighboring repeaters accessible by the new repeater have collected all the route information broadcasted by the neighboring repeaters. Based on the route information collected from neighboring repeaters, the path cost is calculated and a provisional parent repeater is selected.

2 shows an example of an MMR network. In the illustrated MMR network, there are three repeaters (RS1, RS2, RS3) for relaying data in three base stations (BS1, BS2, BS3) and a base station and a new RS have. This new repeater periodically receives UCD messages from each neighboring repeater. Based on the path information of the received message, the new repeater calculates the path cost to each base station (BS1, BS2, BS3) to generate the path metrics. Finally, RS1, RS2 and RS3 select the optimal parent repeater for communication.

3 summarizes the message flow required for the initial network entry procedure in the MMR network shown in FIG. First, a new RS generates a path metrics table as soon as it receives a UCD message. And continuously receives UCD messages from other base stations. As soon as the new repeater receives the already received UCD message by periodic broadcasting, it ends the generation of the path metrics table. In FIG. 2, if the relay RS2 has the lowest path cost, the new relay 'new RS' selects BS2 connected to RS2 as the serving base station.

When selecting a path, both the loss rate and the link bandwidth should be considered. Since the data throughput of a wireless link depends on the bandwidth of the link and the loss rate of the physical layer, it is desirable to consider both of them in path selection.

The hop-count is also considered in the path calculation. This is due to two reasons. First, as the hop-count of the repeater increases, more radio resources are consumed in the relay communication. Also, the increasing hop-count causes the overall delay of the path to increase, thereby degrading network performance. The increased number of hops in a TCP connection increases the round trip delay (RTT) and thus reduces the data throughput. Therefore, it is desirable to minimize the influence of the hop-number by selecting a path with as few hops as possible.

The path metrics that take into account the above mentioned requirements are as follows. In the metrics proposed according to the present invention, the link available bandwidth, MCS level, hop-number, etc. are considered. In the following description, B _xy denotes an available bandwidth between nodes x and y, M _xy denotes an MCS (Modulation Coding Scheme) level, and C _xy denotes a hop-number.

Link available bandwidth: First, it is assumed that most routing paths are derived from the available ratios of resources such as bandwidth. The available bandwidth means the remaining bandwidth not used in the entire bandwidth allocated to the link, where B _{r -Bsi} denotes the set of repeater link bandwidths b from the new repeater r to the base station BS _i .

Here, j ₁ , j ₂ , ... j _k denotes intermediate repeaters between the new repeater and the base station.

MCS link adaptation: The MCS level suitable for the radio channel condition at each instant has the highest average data rate. The MCS level applied according to the state of the radio channel is also referred to as a link adaptation process, and is a method of applying modulation according to the channel state of the link.

Table 1 is a typical sample applied according to the magnitude of the SNR threshold. M _r-BS is a set of MCS levels of each link between repeater r and base station BS _i .

Here, j is defined as an intermediate repeater as shown in equation (1).

Hop count : The hop count is used to measure the path cost of the network. A path with a small number of hops consumes less radio resources of the network and has relatively low noise. Therefore, we prefer a small number of hops when selecting a route. This value is accumulated from the base station to each repeater. The number of hops from the repeater r to the base station BS _i is denoted by H _{r - BS i} .

Consider the above three requirements and calculate the path cost from the repeater to the base station. The ideal path is a path with less hop-count and less robust channel coding scheme that balances the load balance with other paths.

As a result, the data throughput of each path link is expressed as the product of b _xy (the bandwidth of the repeaters x and y) and m _xy (the MCS level of the repeaters x and y), and L _{r -B BSi} is expressed as Defining the smallest link data throughput among the set of links, it is expressed as follows.

When selecting a route, the new repeater enumerates all candidate repeater paths from 1 to n. Where n is the number of routable repeaters. The data throughput L _{r -Bsi} (n) (n-th path from the relay r to the i-th base station BS _i ) of the link having the smallest expected data throughput rate in each path up to one base station is found. And because the path of a small number of hops preferred impart the _{L-BSi r} (n) to calculate a route to the function H _r-BS (n).

As a result, the largest path cost C _{r -Bsi} among the values of the path data throughputs P _{r -Bsi} (n) from the candidate paths 1 to n from the relay r to the base station BS _i is obtained.

This path is an optimal path for deriving the largest data throughput from the newly entering repeater to one base station BS _i . Finally, C _{r - BS i} is calculated at each base station BS _i , and the path having the largest value is selected as an optimal path among the entire network paths. Therefore, the metrics proposed in this paper can select the path between the repeater and the base station with the highest data throughput that can be achieved by eliminating the bottleneck link.

For ease of understanding, based on the example of FIG. 2, FIG. 4 shows the metrics and simple path table flow proposed in accordance with the present invention. In the figure, each repeater obtains path information through a parameter of a UCD message broadcast from a neighboring repeater. As shown in FIG. 4, each repeater completes the path table and selects the path having the smallest path cost. After selecting the path, the repeater periodically broadcasts the path parameter value to the child RS. This sequence of steps is repeated until the new RS obtains the path metrics.

One important consideration is how often this action is repeated. Ideally, if the repeater knows the available bandwidth information of all the links, the repeater will continuously update and determine the most recent information. However, this method is not scalable or practical because it is almost always available bandwidth with frequent updates.

Thus, in order to compensate for the problem, we promise to update the advertised path when the path cost C _{r - BS i} is advertised twice the previous update value, 2 · C _{r - BS i} . If the current path value decreases to a value less than C _{r - BSi} / 2, the repeater scans the channel again, updates the path table, and starts the path selection process.

FIG. 5 shows three base stations BS1, BS2 and BS3, four repeaters RS1, RS2, RS3 and RS4, and a new repeater RS (new RS). A new repeater is assumed to be located in the zone a, b, c. Estimated data throughput will be very different depending on each location.

Table 2 shows the link SNRs of the new repeater and each neighboring relay RS1, RS2, and RS3. If the new repeater chooses the path to the base station, it will choose RS2, RS1, RS3 as the parent repeater in each zone a, b, c if only considering the link status with the adjacent candidate parent repeater. However, the repeater must consider all repeater link states in the path to the base station. The proposed scheme supports a new repeater to consider all link information from neighboring repeaters to updated base stations. Thus improving the performance of the entire network.

FIG. 6 schematically illustrates the data throughput according to the communication with each base station from the scenario of FIG. The data throughput shown in the graph is a value considering the link state and the number of hops shown in the scenario of FIG.

In Fig. 6, the best parent repeater changes depending on which of a, b, and c is located. When the new repeater is located at a, RS2 and b communicate with RS3 and when c, with RS1. This is different from the previous one in which only the state of the wireless link with which the new repeater directly communicates is considered.

In FIG. 6, the repeater in the zone c does not obtain the path information connected to the BS2 from any adjacent neighbors of the New RS because the neighbors in the area of the new repeater are not connected to the BS2 so that the new repeater never obtains the path information to BS2 Can not.

As described above, according to the present invention, a new method of selecting a route of a repeater in a non-transparent mode of an IEEE 802.16j MMR network is proposed. The repeater can select an optimal path to the base station by the path selection method proposed by the present invention. Also, the metrics used in the proposed algorithm make it easy to select the optimal path. The algorithm proposed in accordance with the present invention has the advantage of being able to select a repeater path that can yield the lowest delay and the highest data throughput.

The path selection method in a distributed manner is more complicated because the repeater must perform similar operations to the base station. Nevertheless, the proposed method according to the present invention uses only a few metrics and utilizes the MAC management message used in the existing IEEE 802.16, so that it is possible to have similar complexity to the network entry using the centralized method .

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments. The technical idea of the present invention should not be construed as being limited.

1 is a schematic diagram illustrating a medium access message configuration used in a path selection method for distributed scheduling of a mobile multi-hop relay (MMR) network in accordance with an embodiment of the present invention;

FIG. 2 is a schematic diagram illustrating an MMR network in which a path selection method for distributed scheduling according to an embodiment of the present invention is used; FIG.

FIG. 3 is a flow chart illustrating a procedure for a new repeater to initiate an initial network in an MMR network using a path selection method for distributed scheduling according to an embodiment of the present invention; FIG.

4 is a schematic diagram illustrating a path cost table that may be generated based on an illustrative D-MMR network in FIG. 2 using a path selection method for distributed scheduling in accordance with an embodiment of the present invention;

5 is a schematic diagram illustrating a path cost table that may be generated in a path selection method for distributed scheduling according to an embodiment of the present invention;

FIG. 6 is a graph illustrating a data throughput of each path according to a position of a new repeater in a path selection method for distributed scheduling according to an embodiment of the present invention. FIG.

DESCRIPTION OF THE REFERENCE NUMERALS

1: UCD MAC message

10: Path parameters

Claims (4)

1. A path selection method for distributed scheduling of a mobile multi-hop relay (MMR) network for a broadband wireless access comprising a plurality of relays (RSs) Each repeater (RS) broadcasting a medium access control message having a parameter value for path selection; A new RS (New RS) entering the MMR network scans the broadcast medium access control message and generates a path metrics table based on the parameter values for the path selection among the received media access control messages; Calculating a path cost for each of the paths between the new RS (New RS) and the base station (BS) in the MMR network based on information in the path metrics table; Selects a repeater of the route having the most optimal cost among the calculated path costs as a parent relay of the new RS and transmits a BS connected to the parent relay to a serving BS of the new RS Select the path selection step Hop relay network according to claim 1, wherein the path selection method comprises the steps of: 2. The method of claim 1, wherein the mobile multi-hop relay (MMR) network for the broadband wireless access is based on IEEE 802.16j, the broadcast medium access control message is a UCD (Uplink Channel Descriptor) Wherein the parameter values for path selection are coded in a Type-Length-Value (TLV) format. 2. The method of claim 1, wherein, in calculating the path cost for each path, the path cost (P) for each path is calculated by dividing the data throughput (L) of each path by the number of hops Is defined as P = L / H, i.e. P = L / H. The method of claim 1, wherein after the path selection step, the new RS periodically broadcasts the path parameter value to a child RS. Path selection method.

Images