KR101468694B1 - Network coding and resource allocation method for multichannel multiradio multirate multisession wireless multihop network - Google Patents

Abstract

The present invention relates to a network coding and resource allocation method in a multichannel multiradio multirate multisession wireless multihop network, comprising a first step of configuring a network utility optimization framework by using the Lagrange function and obtaining, by a control unit, sum information of dual variables with respect to all the sessions in every node of the network, a second step of allocating channels to nodes, starting from a node having the largest value, on the basis of the sum information of dual variables, a third step of performing channel allocation, while reducing radios one by one until when the number of radios that may can be used in a channel-allocated transmission node reaches 0, and a fourth step of returning to the second step to perform channel allocation to a node having next priority on the basis of the sum information of the dual variables and repeating the third step until when all the nodes in the network are terminated, in the multichannel multiradio multirate multisession wireless multihop network environment.

Description

TECHNICAL FIELD [0001] The present invention relates to network coding and resource allocation in a multi-channel multi-radio multi-rate multi-session wireless multi-hop network,

The present invention relates to a resource allocation method in a multi-channel multi-radio multi-rate multi-session wireless multi-hop network, and more particularly, to a method and apparatus for allocating coded packets over an allocated channel by combining inter- To a network coding and resource allocation method in a multi-channel multi-radio multi-rate multi-session wireless multi-hop network for improving throughput.

In general, wireless multi - hop networks can be widely applied to various fields, and recent studies are continuing to utilize them because they can be installed at low cost due to a decrease in hardware prices for wireless connection.

Conventional wireless networks are aimed at connecting a large number of wireless nodes.

However, the conventional wireless network has a problem that performance of the wireless network may be deteriorated due to inter-channel interference when multiple nodes transmit simultaneously. In addition, the conventional wireless network has a problem that the characteristics of the link can be easily changed according to various environments, and the throughput is restricted.

Therefore, network coding for solving the above problems and improving the efficiency of a wireless channel has been proposed. The network coding is a method for transmitting a packet received by a relay node that transmits a packet to a next hop node by combining the received packets into one packet differently from an existing simple forwarding routing method, Can be efficiently utilized to improve the throughput of the network (i.e., the degree of how fast data can be transmitted).

The network coding is roughly classified into two types. The first is inter-session network coding, which is a method of combining packets of different flows, and the second is intra-session network coding, which combines packets of the same flow.

The inter-session network coding can carry information in a smaller number of packet transmissions than in an overhearing network structure such as an X-type (X-topology). However, the inter-session network coding is greatly limited by the loss rate of the link. For example, at a loss rate of more than 20%, the inter-session network coding does not function properly and the throughput is significantly lowered.

A high transmission rate can be used as one method for improving the throughput. However, when the same power is used, the transmission speed and the transmission radius are in a trade off relationship. Therefore, transmission at a high transmission rate reduces the transmission range and causes a performance degradation. In particular, since a single rate environment has one optimum point between two conditions (for example, transmission rate and transmission range), it can not cope with a rapidly changing wireless network situation and can lead to a drop in performance due to a decrease in transmission probability. On the other hand, since the multi-rate environment has various transmission speeds, it is possible to select a proper transmission rate according to the wireless network situation and to prevent the decrease in the transmission probability. To do this, it is necessary to select an optimum point among the appropriate transmission rate and transmission range.

Further, research has been conducted on using multiple channels / multiple radios as a method (or a method for improving the throughput) that can solve the decrease in throughput due to co-channel interference and channel competition between nodes in a wireless network. That is, in the multi-channel / multi-radio environment, nodes can use different orthogonal channels using a plurality of radios, so that simultaneous transmission between nodes can be performed, thereby increasing the throughput. This requires sophisticated resource allocation algorithms (ie, for efficient network coding in multi-channel / multi-radio environments).

Also, as a method for minimizing performance degradation due to transmission loss in a wireless network, a congestion control scheme considering the network coding is used.

The congestion control scheme is a scheme that utilizes the size of a queue of each node indicating congestion through inter-node feedback in the transmission process. For example, if congestion occurs during transmission, the source node adjusts the number of transmission packets to minimize the performance degradation caused by congestion. However, the conventional congestion control scheme considering network coding is limited to a single channel / single radio network environment. Therefore, the conventional congestion control scheme considering network coding has a problem that it can not be applied to a multi-channel / multi-radio network environment which is being utilized recently.

Therefore, there is a need for a method for optimizing throughput (that is, throughput improvement) in multi-channel / multi-radio / multi-rate / multi-session wireless multi-hop networks,

BACKGROUND ART [0002] The background art of the present invention is disclosed in Korean Patent Laid-Open Publication No. 2010-0068168 (published on Jun.22, 2010, entitled " Method and Apparatus for Data Transfer in Multi-hop Wireless Network).

The present invention has been made to solve the above problems, and it is an object of the present invention to provide a multi-channel multi-radio multi-rate multi-channel multi- And a network coding and resource allocation method in a session wireless multi-hop network.

A network coding and resource allocation method in a multi-channel multi-channel multi-session multi-hop wireless multi-hop network according to an aspect of the present invention includes a network utility optimization framework in a multi-channel multi- A first step of constructing using a Lagrandian dual function, and a controller obtaining sum information of dual variables for all sessions in all nodes on the network; A second step of starting channel allocation from a node having a largest value based on the sum information of the dual variables; A third step of performing channel allocation by decreasing the radio by one until the number of available radios in the transmission node to which the channel is allocated becomes 0; And a fourth step of returning to the second step until all nodes on the network are terminated, performing channel assignment for the next priority node based on the sum of the dual variables, and repeating the third step .

In the present invention, the dual variable is a congestion price of each node considering a channel capacity.

The present invention is characterized in that a flow rate is updated in a transmission link to which a channel is assigned according to a result of channel assignment, and a flow rate is 0 in a link to which a channel is not assigned.

The present invention is characterized in that a channel assignment is performed prior to a session selected according to a throughput calculated based on a link congestion, a broadcast rate, a transmission rate, and a transmission probability.

In the present invention, when the node is an intermediate transfer node, the control unit controls the network coding unit to combine the packets of the same session received from the source node using intra-session network coding, and transmits a parity packet And transmitting the intra-session network coded packets of different sessions to the assigned channel by inter-session network coding.

A network coding and resource allocation method in a multi-channel multi-channel multi-session multi-hop wireless multi-hop network according to another aspect of the present invention includes a network utility optimization framework in a multi-channel multi- A first step of calculating a flow rate of each session source node based on Lagrange multipliers and variables set in advance by a controller, using a Lagrangian dual function; A second step of calculating a flow rate of all nodes on the network; A third step of calculating a broadcast rate on the network; Calculating sum information of dual variables for all sessions in all the nodes on the network and comparing the calculated values with each other; allocating a channel from a node having a largest result value obtained by comparing the sum information of the dual variables; Step 5; And a sixth step of repeating the first through fifth steps.

In the present invention, the utility maximization problem of the network utility optimization framework configured by using the Lagrandian dual function is expressed by three sub-problems , And each of the three sub-problems is calculated through a process of optimizing the three sub-problems.

In the present invention, the three sub-problems are characterized by a problem with congestion control, a routing problem, and a broadcast rate adjustment problem.

In the present invention, the problem of the congestion control is that the packet amount of the link unit transmitted from each node is replaced with the packet amount of each node, and the result of the packet transmission of the source node and the destination node is reflected to calculate the problem solution .

In the present invention, the routing problem is characterized in that the number of packets is determined so that the number of packets to be transmitted from the intermediate node becomes maximum based on the difference in traffic between the transmitting node and the receiving node and the cost of the link, .

In the present invention, the three sub-problems are based on the congestion price of each node considering a coefficient indicating whether a packet transmitted from a source node reaches a destination node of the transmission session, a cost for a job, and a channel capacity .

In the present invention, when the node is an intermediate transfer node, the control unit controls the network coding unit to combine the packets of the same session received from the source node using intra-session network coding, and transmits a parity packet And transmitting the intra-session network coded packets of different sessions to the assigned channel by inter-session network coding.

The present invention combines inter-session network coding and intra-session network coding in a multi-channel multi-radio multi-rate multi-session wireless multihop network to improve the throughput by transmitting coded packets to a destination node over an assigned channel.

FIG. 1 is a diagram illustrating a schematic configuration of a resource allocation apparatus in a multi-channel multi-radio multi-rate multi-hop wireless multi-hop network according to an embodiment of the present invention. FIG.
BACKGROUND OF THE INVENTION 1. Field of the Invention [0001] The present invention relates to a method for transmitting a packet in a wireless network including multiple-session multiple nodes in an X-topology environment according to an embodiment of the present invention.
FIG. 3 is a graph illustrating a throughput of a wireless network having six sessions in a grid topology environment as a result of performance evaluation of a method according to an exemplary embodiment of the present invention. FIG. ≪ / RTI > FIG.
5 is a graph illustrating a throughput per channel number according to a change in the number of radios in a wireless network having six sessions in a grid topology environment as a result of performance evaluation according to an embodiment of the present invention. FIG. 4 is a graph illustrating throughputs of the number of channels according to a change in the number of radio in a wireless network having six sessions in a random topology environment. FIG.
FIG. 7 is a graph illustrating the performance of the method according to an exemplary embodiment of the present invention. FIG. 8 is a graph illustrating throughputs per channel and number of radio according to the number of sessions in a grid topology environment. FIG. 8 is a graph showing throughput rates of the number of channels and radio according to the number of changes. FIG.
FIG. 9 is a graph illustrating the throughput per link loss according to a network coding scheme in a grid topology environment, and FIG. 10 is a graph illustrating network coding according to a network topology in a random topology environment. FIG. 5 is a graph illustrating a throughput rate of a link according to a technique; FIG.

Hereinafter, a method of network coding and resource allocation in a multi-channel multi-radio multi-session multi-hop wireless multi-hop network according to the present invention will be described with reference to the accompanying drawings.

In this process, the thicknesses of the lines and the sizes of the components shown in the drawings may be exaggerated for clarity and convenience of explanation. In addition, the terms described below are defined in consideration of the functions of the present invention, which may vary depending on the intention or custom of the user, the operator. Therefore, definitions of these terms should be made based on the contents throughout this specification.

The present invention relates to an apparatus and method for allocating resources for improving throughput in a multi-channel multi-radio multi-rate multi-session wireless multihop network by combining inter-session network coding and intra-session network coding.

For example, the present invention utilizes intra-session network coding at an intermediate delivery node to combine packets of the same session and generate parity packets to compensate for link loss. The inter-session network coded packets of the different sessions coded by the intra-session network are transmitted.

By using a method combining intra-session network coding and inter-session network coding as described above, the present invention solves the problem that the throughput due to packet loss is reduced even with a loss rate exceeding 20%, and achieves the effect of inter-session network coding . In addition, since the present invention does not require a process of selecting a packet to be transmitted by using intra-session network coding, scheduling is simplified and the throughput is improved.

Hereinafter, it is assumed that the base station according to the present embodiment has a multi-channel / multi-radio / multi-rate / multi-session wireless network.

로 나타낸다. 그리고 전체 네트워크에서 사용 가능한 직교 채널의 수를 K로 나타내며

의 조건을 갖는다. 그리고 무선 네트워크를 구성하는(즉, 패킷 전송을 위해 네트워크상의 노드를 공유하여 사용하는) 유니캐스트 세션들의 집합을

로 나타내며, 하나의 세션 a에 대하여 데이터(즉, 패킷)를 생성해 전송을 시작하는 소스 노드(Source Node)와 데이터(즉, 패킷) 전달의 최종 목적지(즉, 종착지)인 목적지 노드(Target Node)의 쌍을

로 나타낸다. 또한 상기 소스 노드에서 다음 홉 노드들로 전송되는 세션 a에 대한 플로우 레이트는

이며 세션 a의 유틸리티 함수를

으로 나타낸다. 상기 유틸리티 함수는 각 세션 a의 소스 노드(

)에서 전송하는 플로우 레이트에 대한 단조 증가 및 강오목(strictly concave) 함수이다. Here, in the network consisting of N nodes, the number of radio stations that each node (i)

Respectively. The number of orthogonal channels available in the entire network is denoted by K

. And a set of unicast sessions that make up the wireless network (i.e., share nodes on the network for packet transmission)

(Source Node) that generates data (i.e., packets) for one session a and starts transmission, and a destination node (Target Node) that is a final destination ) Pair

Respectively. The flow rate for the session a transmitted from the source node to the next hop nodes is

And a utility function of session a

Respectively. The utility function includes a source node < RTI ID = 0.0 >

) Is a monotone increase and a strictly concave function for the flow rate to be transmitted.

At this time, for each unicast session, the remaining nodes except for the source node and the destination node become intermediate nodes and transmit data (i.e., packets). A plurality of sessions share data in the same network environment and transmit data (i.e., packets).

Here, R represents the set of data rates that each node can select.

는 노드 i의 전송 속도를 나타내며, 또한 데이터 레이트의 집합

의 원소 중 하나를 선택한다.for example

Represents the transmission rate of node i, and also represents a set of data rates

Is selected.

로 나타낸다.And topology graph of a multi-hop wireless network

Respectively.

이 된다. Where V represents the set of transport nodes

.

로 나타낼 때

가 된다. 그리고 각 노드의 배치는 고정되어 있다고 가정한다. 또한 채널 k를 사용하는 세션 a의 링크

의 플로우 레이트를

로 나타낸다. The set of transport links is denoted by E, and the link between node i and node j

When expressed as

. It is assumed that the arrangement of each node is fixed. Also, the link of session a using channel k

The flow rate of

Respectively.

는 직접연결 링크의 전달확률을 나타내고,

는 오버히어링 링크의 전달확률을 의미한다. In this embodiment, the probability of link transmission is expressed by the following two methods. That is, in a node where inter-session network coding takes place

Represents the transfer probability of the direct connection link,

Is the probability of overhearing link transmission.

은 중요한 의미를 갖는다. Here, session a is coded with another session a '. If the overhearing link's probability of delivery is low, the packet of session a 'is not delivered and it is difficult to decode the packet of session a at the next hop node,

Have an important meaning.

Utilizing the set network variables, a utility maximization framework of the network can be constructed as shown in the following Equation (1).

는 각 세션 a의 소스 노드에서 전송되는 플로우 레이트를 나타낸다. In Equation (1), (1) denotes an objective function for obtaining the maximum throughput of the network,

Represents the flow rate transmitted from the source node of each session a.

In Equations (1), (2) to (7) are conditional expressions for the objective function (1). (2) means that the flow rate received from one node is equal to the flow rate transmitted. That is, it indicates that all the packets transmitted from the source node reach the destination node, indicating that the destination node can acquire the information (i.e., source information) transmitted from the source node through decoding in the transmission of the intra-session network coding scheme .

In Equation (1), (3) denotes the flow rate condition of the link, and this value indicates that the direction does not indicate directionality.

In Equation (1), (4) means that the number of packets that can be transmitted at a node where inter-session network coding occurs is limited by the broadcast rate and the transmission rate.

The first term relates to packets that are lost in a direct delivery link.

는 세션 a의 패킷 중 인터세션 네트워크 코딩하여 전송할 양을 나타내고, 전달확률

에 의해 그 양이 조절된다. 만약 전송 링크에서 손실이 생기면 전달확률

에 따라 패리티 패킷을 생성하여 손실이 생길 수 있는 패킷의 수를 예측하여 그만큼 더 전송하게 된다. 여기서

는 데이터와 리던던시를 포함한 세션 a의 총 레이트를 의미한다. In other words,

Indicates the amount to be transmitted in the inter-session network coding among the packets of the session a,

The amount of which is controlled by the controller. If there is a loss in the transmission link,

A parity packet is generated according to the number of packets that can be lost, and the number of packets is predicted. here

Means the total rate of session a, including data and redundancy.

And the second term relates to the loss of the overhearing link.

는 세션 a'의 패킷에 대한 손실을 보상하기 위해 노드 i에서 인트라세션을 이용해 생성한 리던던시이다. 오버히어링 링크에서 손실된 a'의 패킷 수만큼 중간 전달 노드에서 추가적으로 전송함으로써 세션 a의 디코딩이 가능하게 된다. 두 번째 항 역시 직접 전달 링크의 손실을 고려하여 패킷의 양을 추가하여 전송하게 된다. In other words,

Is a redundancy created using an intra session at node i to compensate for the loss of the packet of session a '. The additional transmission at the intermediate transfer node by the number of packets a 'lost in the overhearing link enables decoding of the session a. The second term is also transmitted by adding the amount of the packet considering the loss of the direct transmission link.

는 세션 a'에 대한 리던던시의 총 레이트이다. 분모에

를 추가함으로써 직접 전달 링크에서 손실되는 패킷을 중간 전달 노드에서 보상함으로써 패킷의 전송 손실을 최소화한다. here

Is the total rate of redundancy for session a '. In the denominator

To compensate for the packets lost in the direct delivery link at the intermediate delivery node, thereby minimizing the packet transmission loss.

는 인터세션 네트워크 코딩이 일어나는 노드에서 네트워크 코딩을 진행하는 비율을 의미한다. 즉, 어떤 세션의 패킷들이 특정 노드 i에 수신되면

의 크기에 따라 인터세션 네트워크 코딩을 진행한다. 상기 (5)는 인터세션 네트워크 코딩이 일어나는 노드가 수신한 패킷을 모두 다음 홉 노드로 전송해야 함을 의미한다. 즉, 해당 노드 i에 진입한 패킷들은

크기에 비례하여 인터세션 네트워크 코딩되어 전송되고, 나머지 패킷들은 코딩되지 않은 채 다음 홉 노드 j로 전송됨을 의미한다. Further, in (5) of Equation 1,

Quot; refers to the rate at which network coding proceeds at a node where inter-session network coding takes place. That is, when packets of a certain session are received at a specific node i

The inter-session network coding is performed according to the size of the inter-session network. (5) means that all the packets received by the node in which the inter-session network coding occurs should be transmitted to the next hop node. That is, the packets entering the node i

And the remaining packets are transmitted to the next hop node j without being coded.

는 노드 i가 세션 a의 패킷에 대해 채널 k를 이용하여 전송할 때 브로드캐스트 레이트를 나타낸다. 또한

는 채널 k의 용량을 나타낸다. 즉, 노드 i와 이 노드의 전송 반경 내의 노드 j는 세션 a에 대하여 채널 k의 용량을 나누어 쓰고 있음을 의미하며, 브로드캐스트 레이트

는 채널용량에 의해 제한을 받는다. In Equations (1) through (6), channel scheduling conditions are shown in the network. here,

Represents the broadcast rate when node i transmits a packet of session a using channel k. Also

Represents the capacity of channel k. That is, the node i and the node j within the transmission radius of this node mean that the capacity of the channel k is divided for the session a, and the broadcast rate

Is limited by the channel capacity.

In Equation (1), (7) is a flow rate condition expression expressed by a broadcast rate, a data rate, and a delivery probability. That is, the flow rate of the link indicates that the transmission rate and the transmission probability of the link should be considered.

The base environment of the wireless network having the multi-channel / multi-radio / multi-rate / multi-session according to the present embodiment has been described above.

1 is a diagram illustrating a schematic configuration of a resource allocation apparatus in a multi-channel multi-radio multi-rate multi-session wireless multi-hop network according to an embodiment of the present invention.

The resource allocation apparatus 100 in the multi-channel multi-channel multi-session multi-hop wireless multi-hop network according to an exemplary embodiment of the present invention includes a packet receiver 110, a storage unit 120, a controller 130, (140), and a packet transmission unit (150).

또는 목적지 노드

가 될 수 있다.The resource allocation apparatus 100 may configure a transport node constituting a multi-channel multi-radio multi-rate multi-session wireless multi-hop network,

Or destination node

.

The packet receiving unit 110 receives the resource allocation apparatus 100 as a receiving node in a multi-channel multi-radio multi-rate multi-session wireless multi-hop network, . In addition, the packet receiving unit 110 may simultaneously receive a packet through a plurality of links formed by using a plurality of different nodes, each of which is assigned a different channel, as a transmitting node.

The storage unit 120 stores the packets received by the packet receiving unit 110.

The mixed network coding unit 140 performs the intra-session network coding and the inter-session network coding on the packets stored in the storage unit 120.

For example, when the resource allocation apparatus 100 operates as an intermediate transfer node, the control unit 130 controls the mixed network coding unit 140 to transmit packets of the same session received through the packet receiving unit 110 to the intra- Session network coding, parity packets to compensate for link loss, and inter-session network coding of the intra-session network coded packets of different sessions.

Accordingly, the mixed network coding unit 140 may include an intra-session coding unit 141 for intra-session network coding and an inter-session coding unit 142 for inter-session network coding.

The control unit 130 transmits the intra-session network coding and inter-session network coded packets to the destination node through the packet transmission unit 150.

The packet transmission unit 150 transmits a mixed network coding packet to the mixed network coding unit 140 through a channel allocated by the resource allocation apparatus 100.

The control unit 130 may control operations of the packet receiving unit 110, the storage unit 120, the mixed network coding unit 140, and the packet transmission unit 150.

일 경우 후술하는 본 발명의 일 실시예에 따른 다중채널 다중라디오 다중레이트 다중세션 무선 다중홉 네트워크에서의 자원 할당 방법에 따라 소스 노드

의 플로우 레이트(

)를 산출할 수 있다.In addition, if the resource allocation apparatus 100 determines that the source node

In accordance with a resource allocation method in a multi-channel multi-radio multi-session multi-hop wireless multi-hop network according to an embodiment of the present invention,

Of the flow rate

) Can be calculated.

)를 산출할 수 있다.In addition, the controller 130 controls the resource allocation apparatus 100 according to a resource allocation method in a multi-channel multi-radio multiple-rate multi-session wireless multi-hop network according to an embodiment of the present invention, Of the flow rate

) Can be calculated.

In addition, the controller 130 calculates a broadcast rate of the resource allocation apparatus 100 according to a resource allocation method in a multi-channel multi-radio multiple-rate multi-session wireless multi-hop network according to an embodiment of the present invention, A channel can be allocated to the allocating apparatus 100. [

Hereinafter, a resource allocation method according to an embodiment of the present invention will be described.

In this embodiment, a method for selecting an efficient transmission rate in a multi-rate environment, an algorithm for modifying the flow rate and a broadcast rate by formally solving the utility maximization framework, and a multi-channel / multi-radio wireless network Provides a resource allocation algorithm for full use of available bandwidth.

First, a method of selecting a transmission rate in the multi-rate environment will be described. That is, a method of selecting a transmission rate of each node constituting a wireless network using a multi-rate will be described.

로 나타내기로 한다. 그리고 노드 i가 전송 속도

로 전송할 때, 노드 i의 전송 범위를

이라고 한다. 그리고 링크의 길이를

로 나타낸다. 이에 따라

의 조건을 만족하는 노드를 전송 후보 노드라고 하며, 이러한 노드의 집합을

로 나타낸다. 그리고 노드 i와 집합

사이에서 ETR(Effective Transmission Rate)을 아래의 수학식 2와 같이 정의한다. Hereinafter, a set of available transmission rates at the multiple rates

. Then,

, The transmission range of node i is

. And the length of the link

Respectively. Accordingly

The node that satisfies the condition of < RTI ID = 0.0 >

Respectively. And node i and set

ETR (Effective Transmission Rate) is defined as Equation (2) below.

Here, the effective transmission rate (ETR) value means the number of bits for which a valid transmission from the current transmission node to the destination node is expected.

는 노드 i에서 전송 속도

로 전송했을 경우의 ETR 값을 의미하고, 우변의

는 노드 i와 j 사이의 전달확률이다. 그리고

는 노드 i의 전송 후보 노드 중 하나인 노드 j의 ETR(Effective Transmission Rate) 값을 나타낸다. Here,

Lt; RTI ID = 0.0 > i &

Quot; ETR " means "

Is the transfer probability between nodes i and j. And

Represents the effective transmission rate (ETR) value of node j, which is one of the transmission candidate nodes of node i.

A method of calculating the Effective Transmission Rate (ETR) and selecting a transmission rate of each node will be described with reference to Table 1 below.

에 대하여 수학식 1의 (7) 계산.
3) 노드 i의 ETR 중 가장 큰 값을 갖도록 하는 전송 속도

을 전송 속도로 선택.
4) 소스 노드의 전송 속도 선택이 완료될 때까지 상기 1)~3)의 과정을 반복.1) Initialize ETR value of all nodes to 0.
2) from the node closest to the destination node of the session

(7) in Equation (1) with respect to Equation (1).
3) a transmission rate at which the ETR of node i has the largest value

As the transmission speed.
4) Repeat the above steps 1) to 3) until selection of the transmission rate of the source node is completed.

Since the effective transmission rate is calculated from the node nearest to the destination node, the Effective Transmission Rate (ETR) value of the transmission candidate node is included in the calculation.

The transmission rate selection process described with reference to Table 1 can be summarized as Equation 3 below.

An algorithm for adjusting the flow rate and the broadcast rate by formally solving the utility optimization framework constructed as shown in Equation (1) will be described below. That is, a dispersion rate adjustment algorithm will be described.

The embodiment of the present invention provides an algorithm that can solve the network utility maximization problem by using a dual decomposition using a Lagrangian dual function.

The above equation (1) can be expressed by the Lagrangian dual function.

Equation (4) is equivalent to the utility maximization problem by Equation (5).

The dual function using the Lagrangian dual function can be expressed by Equation (6) below.

Equation (6) is divided into three subproblems and summarized as Equation (7).

Equation (7) represents a process for optimizing each of three subproblems obtained through decomposition.

은 혼잡제어에 관한 문제를 나타내고,

는 라우팅 문제를 나타내며,

는 브로드캐스트 레이트 조정 문제를 나타낸다. 상기

의 문제는 라그랑주 승수(Lagrange multiplier)

의 값을 활용하여 독립적으로 풀 수 있다. 각각의 라그랑주 승수(Lagrange multiplier)의 값을 서브그래디언트 기법(subgradient method)을 사용하여 구하면 다음 수학식 8과 같다.here

Indicates a problem with congestion control,

Indicates a routing problem,

Indicates a broadcast rate adjustment problem. remind

The Lagrangian multiplier is a problem,

Can be solved independently. The value of each Lagrange multiplier is obtained by using the subgradient method.

는 스텝사이즈(step size)로서 각각 승수(multiplier)의 값을 결정하기 위해 사용되는 변수이다. 상기 각각의 스텝사이즈(step size)를 정하기 위해 환산계수(scaling factor)

의 값을 임의의 상수로 정하고, 상기 스텝사이즈(step size)는

로 구한다. 또한 상기 수학식 8에서

는

의 범위의 값만을 가질 수 있음을 나타낸다.When calculating the respective Lagrange multipliers,

Is a step size, which is a variable used to determine the value of a multiplier, respectively. A scaling factor is used to determine each step size.

Is set to an arbitrary constant, and the step size is set to

. In Equation (8)

The

Lt; RTI ID = 0.0 > of < / RTI >

의 수학식은 소스 노드에서 전송할 패킷량을 계산하는 과정이다.remind

Is a process of calculating the amount of packets to be transmitted from the source node.

의 수학식 문제를 풀기 위하여, 상기 수학식 7에서 상기

의 수학식을 아래의 수학식 9와 같이 변형할 수 있다. 이는 현재 각 노드에서 전송하는 링크 단위의 양을 노드 단위의 양으로 바꾸어 소스 노드와 목적지 노드의 패킷 전송의 결과를 반영하여 혼잡 제어를 반영하기 위한 과정이다. 상기

의 수학식에서

로 정리하면, 수학식 9의 (1)과 같다. 그리고 상기 수학식 9의 (1)을 1차 최적조건(first-oder optimal condition)에 의하여 정리하면 수학식 9의 (2)와 같은 해를 얻을 수 있다.remind

In order to solve the problem of the equation

Can be transformed into the following equation (9). This is a process for reflecting the congestion control by reflecting the result of packet transmission between the source node and the destination node by changing the amount of the link unit transmitted from each node to the amount per node unit. remind

In Equation

(1) of the equation (9). If (1) of the above equation (9) is summarized by the first-order optimal condition, a solution like the equation (2) of the equation (9) can be obtained.

는

와

에 의하여 결정되며, 서로 반비례 관계에 있음을 알 수 있다. 만약 트래픽 흐름이 원활하지 않아 소스 노드에서 패킷이 전송되지 않고 정체될 경우

가 상승하여 소스 노드에서 전송하는 패킷의 양이 감소된다. 따라서

는 소스 노드의 전송이 목적지 노드에 도달하는지 여부를 나타내어 네트워크의 종단간 피드백을 반영함을 알 수 있다.In Equation (9), the utility function uses a log function,

The

Wow

And they are inversely proportional to each other. If the traffic flow is not smooth and the source node is congested

The amount of packets transmitted from the source node is reduced. therefore

Indicates whether the transmission of the source node reaches the destination node and reflects the end-to-end feedback of the network.

의 수학식은 라우팅 결과를 반영하여, 중간 노드에서 전송할 패킷 수를 결정한다. 이때 송신 노드와 수신 노드 사이의 트래픽 차이와 링크의 코스트(cost)를 고려하여 중간 노드에서 전송할 패킷 수가 최대가 되게 한다. 상기

의 수학식에서

는 노드 i에서 세션 a에 해당하는 패킷이 인터세션 네트워크 코딩될 때에 한해서 1이 되는 변수이고, 그 이외의 경우에는 0으로 정의된다. 이때

의 값이 증가하면 백프레셔(back-pressure)가 증가하여 중간 노드가 전송할 패킷 수

가 상승하게 된다. 이때 상기

의 수학식 문제는 작업 스케줄링 문제(job scheduling problem)의 일반화 과정처럼 해석할 수 있으며,

는 작업(job)에 대한 코스트(cost)를 나타낸다.remind

The number of packets to be transmitted from the intermediate node is determined by reflecting the routing result. At this time, considering the traffic difference between the transmitting node and the receiving node and the cost of the link, the number of packets to be transmitted from the intermediate node is maximized. remind

In Equation

Is a variable that is 1 when the packet corresponding to the session a in the node i is inter-session network coded, and is defined as 0 otherwise. At this time

The back-pressure is increased and the number of packets to be transmitted by the intermediate node

. At this time,

Can be interpreted as a generalization process of a job scheduling problem,

Represents the cost for a job.

FIG. 2 is a diagram illustrating an exemplary packet transmission method in a wireless network configured with multiple-session multiple nodes in an X-topology environment according to an exemplary embodiment of the present invention. Referring to FIG.

The X-topology as shown in FIG. 2 includes Session 1 transmitting from node A to node D via node A, and session 2 transmitting from node B to node C via intermediate node I.

Here, the node C and the node D overhear the packets transmitted by the node A and the node B, respectively. And the loss rate of the link between the intermediate node I and the node D and the overhearing link between the node B and the node D is 50%, respectively.

At this time, when the session 1 and the session 2 transmit four packets, respectively, four packets are transmitted to the intermediate node I for each session. Since the intermediate node I has a loss of 50% in the link connected to the node D, 4 additional parity packets of the session 1 are generated using the intra-session network coding. In addition, since a loss occurs in the overhearing link of the session 2, a problem occurs that the node D overhears only two of the four packets of the session 2 and can not restore the packet of the session 1. Therefore, Lt; / RTI > In this case, since there is a loss also in the direct transmission link, four parities are generated considering this loss. On the other hand, since there is no loss in the direction in which the session 2 is transmitted, the parity packet is not generated.

Therefore, the intermediate node I has 12 packets in total for session 1, four packets received from node A, four parities for direct link loss, and four parity for overhearing link loss of session 2. And for Session 2 there are four packets received from the Node B. When the parity generation is completed, inter-session network coding is performed.

As described above, since the packets of the session 1 are four packets of the twelve sessions 2, four inter-session network coded packets are generated and the remaining eight packets are transmitted without being coded. When the packet transmission is performed as described above, the node D overheats two packets at the node B and receives six packets from the intermediate node I. Accordingly, the node D can decode four packets transmitted from the node A using the total of eight packets received.

의 수학식을 계산하기 위하여

로 치환한다. 그리고 듀얼 변수(dual variable)

는 채널 용량을 고려하였을 때 노드 i의 혼잡 가격(congestion price)을 의미한다. 상기 듀얼 변수

의 값은 수학식 8에서 계산할 수 있다. 한편 상기

의 수학식은 선형함수이기 때문에 프리멀(primal)한 해를 가질 수 없다. 따라서 근부 최적화 알고리즘(proximal optimization algorithm)을 사용하여 2차 항을 추가하여 강볼록(strictly convex) 함수로 변형하고 정리하면 다음 수학식 10과 같다.remind

To calculate the equation of

. And dual variables.

Denotes the congestion price of node i when the channel capacity is considered. The dual variable

Can be calculated in Equation (8). Meanwhile,

Can not have a primal solution because it is a linear function. Therefore, the second term is added using the proximal optimization algorithm to transform it into a strictly convex function, and the following equation (10) is obtained.

는 해를 구하기 위해 사용되는 작은 양의 상수이며, 상기 수학식 10을 계산하면 다음 수학식 11과 같이 브로드캐스트 레이트

의 결과를 얻을 수 있다.here

Is a small positive constant used to obtain the solution, and the equation (10) is calculated as follows: < RTI ID = 0.0 >

Can be obtained.

의 선형함수를 변형하여 상기 수학식 11에서 해를 구했기 때문에 계산된 프리멀 가용해(primal feasible solution)가 오차를 가지므로, 프리멀 회복 기법(primal recovery method)을 적용하여 아래의 수학식 12와 같이 프리멀 최적해(primal optimal solution)을 계산할 수 있다.remind

The primal feasible solution calculated because the linear function of equation (11) is transformed and the solution is calculated. Therefore, the primal recovery method is applied to the following equation 12 , We can calculate the primal optimal solution.

A resource allocation algorithm for full use of available bandwidth in a multi-channel / multi-radio wireless network to which a network coding scheme is applied will be described according to an embodiment of the present invention.

The resource allocation algorithm for the multi-channel and multi-radio is based on the assumption that when each node transmits a packet, a channel is allocated to the node link and one radio is used at both ends of the transmitting and receiving nodes.

The original resource allocation process can not be formally solved because the complexity is NP-hard (that is, a set of problems that can be transformed from a potential problem).

(즉, 채널 용량을 고려하였을 때 노드 i의 congestion price)를 활용해 채널 할당의 순서를 결정한다. 이에 따라 혼잡이 발생하는 노드에 우선적으로 채널 할당을 수행하여 패킷 전송의 쏠림을 방지함으로써 네트워크의 처리율을 향상하기 위한 방법(즉, 자원 할당 알고리즘)을 제공한다.Accordingly, in this embodiment, as a heuristic method, a dual variable

(I. E., Congestion price of node i when channel capacity is considered) to determine the order of channel assignment. Accordingly, the present invention provides a method (i.e., a resource allocation algorithm) for improving the throughput of the network by preventing the sinking of the packet transmission by performing the channel allocation preferentially to the node where the congestion occurs.

를 구한다. 그리고 각 노드는 자신을 포함한 이웃들의

정보를 가지게 된다. 상기 정보를 바탕으로 값이 가장 큰 노드부터 채널 할당을 시작한다. 그리고 채널이 할당된 전송의 양 단 노드에서 사용할 수 있는 라디오의 수를 1씩 감소시킨다. 상기 채널 할당 과정은 전송 노드가 더 이상 전송할 곳이 없거나, 사용 가능한 라디오 수가 0이 될 때까지 수행하며, 이 과정이 끝나면 다음 우선순위의 노드로 넘어가 채널 할당이 수행된다. To do this, first of all nodes on the network

. Each node has its own neighbors

Information. Based on the information, the channel allocation is started from the node having the largest value. And decrements the number of radios available on both ends of the channel-assigned transmission by one. The channel allocation process is performed until the transmission node has no more transmission channels or the number of available radio channels becomes zero. When the process is completed, the channel allocation process is performed by going to the next priority node.

가 갱신되지만, 그렇지 않은 링크에서는 플로우 레이트가 0이 된다.The channel allocation continues until the same process is terminated at all nodes on the network. In the transmission link to which the channel is allocated according to the result of the channel assignment,

But the flow rate is 0 in the other link.

The present invention proposes a method of selecting an optimal unicast session.

It is possible to allocate resources to an optimal unicast session considering inter-link traffic congestion, a broadcast rate, a transmission rate, and a transmission probability through the following Equation (13), thereby further improving the throughput.

This paper proposes a network utility maximization problem for multi-channel / multi-radio / multi-rate / multi-session wireless multi-hop network environment and congestion control algorithm and dispersion rate adjustment algorithm based on intra- We explained how to solve the optimization problem by using the resource allocation algorithm.

The above-described process is summarized in Table 2 below.

를 0으로 놓고 y, b를 작은 양의 상수로 설정.
2) 수학식 7의

을 계산하여 각 세션 소스 노드의 플로우 레이트 획득.
수학식 7의

을 계산하여 전체 노드의 플로우 레이트 획득.
3) 수학식 7의

를 계산, 수학식 10에서 브로드캐스트 레이트 획득.
4) 듀얼 변수(dual variable) 계산.
(a) 수학식 8에서

계산.
(b) 수학식 8에서

계산.
(c) 수학식 8에서

계산.
5) 자원 할당 과정 수행.
(a) 각 노드에서

계산 및 주변 노드와 비교.
값이 큰 노드부터 자원 할당 진행.
(b) 수학식 13을 이용하여 최적 세션의 패킷부터 자원 할당.
6) 2)~ 5) 과정 반복.One)

To 0 and set y and b to small positive constants.
2) Equation (7)

To obtain the flow rate of each session source node.
Equation (7)

To obtain the flow rate of the entire node.
3) Equation (7)

0.0 > (10) < / RTI >
4) Dual variable calculation.
(a) In Equation 8,

Calculation.
(b) In Equation 8,

Calculation.
(c) In Equation 8,

Calculation.
5) Performing resource allocation process.
(a) At each node

Computation and comparison with neighboring nodes.
Resource allocation progression from node with large value.
(b) resource allocation from the packet of the optimal session using Equation (13).
6) Repeat steps 2) to 5).

Hereinafter, a performance evaluation result according to an embodiment of the present invention will be described.

A MATLAB program was used for the performance evaluation. First, 16 wireless nodes are assumed, and performance evaluation is performed in a random topology environment and a grid topology environment arranged in a grid form. In this case, each node is composed of multiple radios, and is transmitted through a plurality of channels. The data rates are assumed to be three rates of 1R, 1.5R and 2R, and a single rate situation using only one of three data rates, Are compared with each other. Also, multiple sessions were configured to send packets simultaneously.

First, the throughput change when the algorithm according to this embodiment is repeated is observed and analyzed. The process of maximizing the throughput at this time is shown graphically (see FIGS. 3 to 10). Referring to the graph, it can be seen that the magnitude of the throughput converges constantly and that the optimum solution is obtained in a given situation. We also set up and measure three available radio and channel numbers, assuming an environment in which six unicast sessions operate. It is each (7 radio, 7 channel), (5 radio, 5 channel), (4 radio, 4 channel).

FIG. 3 is a graph illustrating a throughput of a wireless network composed of six sessions in a grid topology environment as a result of performance evaluation of a method according to an exemplary embodiment of the present invention. As a result of the performance evaluation of the method according to the present invention, the throughput of a wireless network composed of six sessions in a random topology environment is shown as a graph.

Referring to FIG. 3 and FIG. 4, it can be seen that the throughput of the network converges in each environment. This indicates that each session is transmitted through the optimal path to the destination node, and also indicates that the resource allocation process is continuously performed at each node. Also, if the same number of sessions are simultaneously transmitted, the higher the number of available channels and radio, the higher the throughput rate is in a multi-rate situation where the available rate is higher than the single rate.

Next, in order to measure the change of the throughput according to the number of available channels and radio, a throughput value obtained by repeatedly repeating the proposed algorithm over a predetermined number of times or more is shown in a graph (see FIGS. 5 and 6). That is, the number of available channels is increased from 4 to 7 in the environment where 6 sessions are simultaneously transmitted, and the result of the network utility is confirmed when the number of radio of each node is fixed to 3, 4, 5.

5 is a graph illustrating a throughput per channel number according to a change in the number of radios in a wireless network having six sessions in a grid topology environment as a result of performance evaluation according to an embodiment of the present invention. Is a graph illustrating a throughput per channel number according to a change in the number of radio in a wireless network having six sessions in a random topology environment as a result of performance evaluation of the method according to an embodiment of the present invention.

Referring to FIG. 5 and FIG. 6, it can be seen that the network utility increases when the number of available channels is large and when many radios can be utilized. Referring to FIG. 5 and FIG. 6, it can be seen that a high throughput can be obtained when a plurality of rates are used.

Next, network utility changes are graphically shown (see FIGS. 7 and 8) when the number of simultaneously transmitting sessions in the network changes. That is, throughput was measured in the environment where the number of channels of each node varied from 4 to 7, and radio was set equal to the number of each channel.

FIG. 7 is a graph illustrating a throughput according to channel and radio numbers according to a change in the number of sessions in a grid topology environment as a result of performance evaluation of a method according to an embodiment of the present invention. FIG. As an example of the performance evaluation result of the method according to the example, graphs are shown in terms of the throughputs per channel and number of radio according to the change in the number of sessions in the random topology environment.

Referring to FIGS. 7 and 8, it can be seen that the greater the number of sessions, the greater the throughput. This shows that the throughput optimization is achieved in multi-channel / multi-radio environments even with a large number of concurrent transmission links, and the throughput is improved by simultaneously transmitting packets of different sessions through inter-session network coding.

Finally, the algorithm according to the present embodiment is shown graphically in comparison with algorithms without inter-session network coding, intra-session network coding, and network coding (see FIGS. 9 and 10). That is, each node uses 7 channels and radio, and the average loss of link is changed by 0 ~ 50%.

FIG. 9 is a graph illustrating a throughput per loss of a link according to a network coding scheme in a grid topology environment as a performance evaluation result of a method according to an embodiment of the present invention. FIG. 4 is a graph illustrating throughputs of link loss according to a network coding scheme in a random topology environment. FIG.

Referring to FIG. 9 and FIG. 10, it can be seen that the throughput is improved when the algorithm according to the present embodiment is used alone when the inter-session or intra-session network is used. In particular, the inter-session network coding can reduce the throughput due to the loss by using the intra-session network coding, while the throughput is severely reduced as the loss increases.

As described above, the present invention relates to a resource allocation method for improving throughput of a multi-channel / multi-radio / multi-rate / multi-session wireless multi-hop network by integrally considering inter-session network coding and intra- The present invention provides a method of using a network coding scheme combining two types of network coding in order to compensate for reduction in throughput due to loss rate without using inter-session network coding and intra-session network coding alone.

That is, the present invention constitutes a problem of maximizing a network utility in a given wireless network environment, and provides a utility maximization method proposed through a process of a multi-rate selection algorithm, a congestion control algorithm, a dispersion rate adjustment algorithm, and a heuristic resource allocation algorithm, I seek the solution of the problem. In addition, through the performance evaluation of the present invention, it is confirmed that the throughput increases when the number of channels and the number of radio increases, and it is confirmed that the utilization of the multi-rate can increase the throughput more than when the single rate is used. It can be seen that an algorithm according to an embodiment of the present invention can contribute to improvement of a wireless network throughput by obtaining a high throughput by combining two types of network coding.

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 embodiments, but, on the contrary, I will understand the point. Accordingly, the technical scope of the present invention should be defined by the following claims.

110: packet receiving unit 120:
130: control unit 140: mixed network coding unit
141: Intra-session coding unit 142: Inter-
150: Packet transmitter

Claims (12)

In a multi-channel multi-radio multi-rate multi-session wireless multi-hop network environment,
A first step of configuring a network utility optimization framework using a Lagrangian dual function, and a controller obtaining sum information of dual variables for all sessions in all nodes on the network;
A second step of starting channel allocation from a node having a largest value based on the sum information of the dual variables;
A third step of performing channel allocation by decreasing the radio by one until the number of available radios in the transmission node to which the channel is allocated becomes 0; And
Returning to the second step until all the nodes on the network are terminated, performing channel assignment for the next priority node based on the sum of the dual variables, and repeating the third step Channel multi-hop wireless multi-hop wireless multi-hop network.
2. The method of claim 1,
Channel multi-hop wireless multi-hop network in a multi-channel multi-radio multi-rate multi-hop wireless multi-hop network.
The method according to claim 1,
Wherein a flow rate is updated in a transmission link to which a channel is assigned according to a result of the channel assignment and a flow rate is 0 in a link to which a channel is not assigned. / RTI >
The method according to claim 1,
Channel multi-hop wireless multi-session multi-hop wireless multi-hop network, wherein the multi-channel multi-channel multi-channel multi-hop multi-hop wireless multi-hop network is characterized in that channel assignment is performed in preference to a session selected according to a throughput calculated based on a link congestion, a link rate, / RTI >
The method according to claim 1,
When the node is an intermediate transfer node, the control unit controls the network coding unit to combine packets of the same session received from the source node using intra-session network coding, and generates a parity packet for compensating for link loss, And transmitting the packets of the different sessions coded in the intra-session network to the assigned channel by inter-session network coding. The multi-channel multi-channel multi-session wireless multi- And resource allocation method.
In a multi-channel multi-radio multi-rate multi-session wireless multi-hop network environment,
A first step of constructing a network utility optimization framework using a Lagrangian dual function, and a controller calculating a flow rate of each session source node based on preset Lagrange multipliers and variables;
A second step of calculating a flow rate of all nodes on the network;
A third step of calculating a broadcast rate on the network;
Calculating sum information of dual variables for all sessions at all nodes on the network and comparing the calculated values with each other;
A fifth step of allocating a channel from a node having a largest result value obtained by comparing the sum information of the dual variables; And
And a sixth step of repeating the first through the fifth steps. The network coding and resource allocation method in a multi-channel multi-radio multi-rate multi-session wireless multi-hop network.
The method according to claim 6,
Wherein the flow rate of each session source node, the flow rate of the entire node, and the broadcast rate divide the utility maximization problem of the network utility optimization framework configured using the Lagrandian dual function into three sub-problems, Wherein each of the sub-problems is computed through optimization of three sub-problems.
8. The method of claim 7, wherein the three sub-
The method comprising the steps of: (a) providing a plurality of radio resources for a multi-channel multi-radio multi-hop wireless multi-hop network;
9. The method of claim 8,
Channel multi-channel multi-session wireless communication system according to claim 1, characterized in that the amount of packets transmitted by each node is converted into a packet amount per node, and the result of the packet transmission of the source node and the destination node is reflected, Network coding and resource allocation in multi - hop networks.
9. The method of claim 8,
Wherein the number of packets is determined so as to maximize the number of packets to be transmitted from the intermediate node based on the difference in traffic between the transmitting node and the receiving node and the cost of the link, thereby calculating the problem solution. A network coding and resource allocation method in a hop network.
8. The method of claim 7, wherein the three sub-
And a congestion price of each node considering a channel cost, a cost for a job, and a coefficient indicating whether a packet transmitted from a source node reaches a destination node of the transmission session. A method for network coding and resource allocation in a rate multi-session wireless multi-hop network.
The method according to claim 6,
When the node is an intermediate transfer node, the control unit controls the network coding unit to combine packets of the same session received from the source node using intra-session network coding, and generates a parity packet for compensating for link loss, And transmitting the packets of the different sessions coded in the intra-session network to the assigned channel by inter-session network coding. The multi-channel multi-channel multi-session wireless multi-hop network coding method according to claim 1, And resource allocation method.

Images