JP2009224836A - Radio device and radio resource allocation method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a radio device and a radio resource allocation method capable of radio resource allocation considering the information of a high-order layer, such as QoS requested by an application layer. <P>SOLUTION: The radio device 1 is provided with a control part 11 for determining the order of terminals to allocate radio resources on the basis of an index of each terminal relating to a communication protocol layer higher than a physical layer and a data link layer. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a radio apparatus and a radio resource allocation method.

Conventionally, in a network, a protocol is multi-layered by a seven-layer model of an OSI (Open Systems Interconnection) reference model, and protocol design is performed for each layer (layer). However, this layer division has a problem that flexible protocol operation is not possible.
In order to solve this problem, a technique for optimizing the entire network across the hierarchy divided into seven layers (hereinafter referred to as cross-layer optimization) has been proposed. In particular, a technique for cross-layer optimization between a data link layer and a physical layer is being studied for application to a wireless communication system. This is because, in wireless communication, characteristics change greatly due to the influence of radio propagation environment and interference in the physical layer, so the influence is absorbed by the radio resource allocation function (scheduling function) of the data link layer, which is one layer above. This is because it is effective. For example, in Non-Patent Document 1, radio resource allocation using physical layer / data link layer information in the downlink (link from base station to terminal) of OFDMA (Orthogonal Frequency Division Multiple Access) system. (Scheduling) methods have been proposed. In this method, radio resources are allocated by proportional fairness scheduling based on data link layer queue state information while maintaining system throughput and fairness among users.

FIG. 4 is a flowchart showing a conventional scheduling flow. The following shows the operation of the scheduler in one radio frame n.
First, in step S101, the scheduler updates the allocatable resource amount. That is, the scheduler calculates the amount of resources that can be allocated for data transmission, excluding physical layer retransmission resources and control message resources. In the next step S102, it is determined whether or not the updated allocatable resource amount is zero. If 0, the process is terminated. If not 0, the process proceeds to step S103.

  In the next step S103, the scheduler selects a user who is an allocation candidate. At this time, the scheduler sets a user whose accumulated amount of the data transmission buffer in the data link layer is not 0 as an allocation candidate. The data transmission buffer holds data to be transmitted to the user. In the next step S104, the scheduler determines the allocation order of the allocation candidate users selected based on the scheduler algorithm. In the next step S105, the scheduler determines the resource amount to be allocated to each allocation candidate user and ends the process. Specifically, the scheduler allocates the necessary resource amount for each user in the order of allocation. The resource amount required for each user is calculated based on the accumulated amount of the data transmission buffer and the packet format determined by CIR (Carrier-to-Interference Ratio). The packet format represents frequency utilization efficiency.

“A new cross layer proportional fairness packet scheduling in the downlink of OFDM wireless system”, Proc IEEE ICC, 2007, pp. 5695-5700, June 2007.

However, in the technique described in Non-Patent Document 1, only the optimization that is closed to the lower layer (data link layer / physical layer) is performed, and QoS (Quality of Service) required by the application layer that is the upper layer is performed. There is a problem that it cannot be satisfied. Further, although scheduling is performed based on information of the physical layer and the data link layer, there is a problem that optimization of the parameters of the physical layer cannot be achieved according to the information each time.
The present invention has been made in view of the above points, and an object of the present invention is to provide a radio apparatus and a radio resource allocation method capable of performing radio resource allocation in consideration of higher layer information such as QoS required by an application layer. It is to provide.

  The present invention has been made in order to solve the above-described problems, and one aspect of the present invention is a wireless device that allocates downlink radio resources to a terminal in a communication protocol layer higher than the physical layer and the data link layer. The wireless apparatus includes a control unit that determines the order of terminals to which wireless resources are allocated based on the index for each terminal.

  One embodiment of the present invention is characterized in that, in the above wireless device, the control unit normalizes each index in accordance with the importance.

  One embodiment of the present invention is characterized in that, in the above wireless device, the index is a QoS class of an application layer or a segment average throughput of a transport layer.

  One embodiment of the present invention is characterized in that, in the above wireless device, the control unit feeds back an optimum parameter for each layer based on an index for each terminal.

  One embodiment of the present invention is characterized in that, in the above wireless device, the control unit feeds back an optimum parameter to each layer based on a QoS class of an application layer.

  One embodiment of the present invention is characterized in that, in the above wireless device, the control unit feeds back a CIR back-off, a maximum number of retransmissions, or a physical layer retransmission delay parameter to the physical layer.

  In addition, according to an aspect of the present invention, in the above wireless device, the control unit feeds back a parameter of the maximum number of retransmissions to a data link layer.

  According to another aspect of the present invention, in a radio resource allocation method for allocating downlink radio resources to terminals, radio resources are allocated based on indices for each terminal related to a communication protocol layer higher than a physical layer and a data link layer. A radio resource allocation method characterized in that the order of terminals is determined.

  One embodiment of the present invention is characterized in that, in the above radio resource allocation method, each index is normalized according to importance.

  In addition, according to one aspect of the present invention, in the radio resource allocation method, the index is a QoS class of an application layer or a segment average throughput of a transport layer.

  Further, one aspect of the present invention is characterized in that, in the above radio resource allocation method, an optimum parameter for each layer is fed back for each terminal based on an index.

  Further, one aspect of the present invention is characterized in that, in the above radio resource allocation method, an optimum parameter is fed back to each layer based on a QoS class of an application layer.

  In addition, according to an aspect of the present invention, in the above radio resource allocation method, a CIR backoff, a maximum number of retransmissions, or a physical layer retransmission delay parameter is fed back to the physical layer.

  Also, one aspect of the present invention is characterized in that, in the above radio resource allocation method, a parameter of the maximum number of retransmissions is fed back to the data link layer.

  According to the present invention, the control unit determines the order of the terminals to which the radio resources are allocated based on the index for each terminal related to the communication protocol layer higher than the physical layer and the data link layer. It can be reflected in resource allocation. As a result, radio resource allocation can be performed in consideration of higher layer information such as QoS required by the application layer.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
FIG. 1 is a block diagram showing a configuration of a wireless device 1 according to an embodiment of the present invention.
The wireless device 1 includes a control unit 11, a wireless unit 12, a network interface 13, and a storage unit 14. The radio | wireless apparatus 1 in this embodiment is comprised in the base station of an OFDMA (Orthogonal Frequency Division Multiplexing; Orthogonal Frequency Division Multiple Access) system. FIG. 1 shows a portion related to the downlink (link from the base station to the terminal). The wireless device 1 transmits data received from the network to the terminal by wireless communication.

  The control unit 11 performs radio resource allocation (scheduling) in the downlink. The network interface 13 receives data to be transmitted to each terminal from the network. The wireless unit 12 transmits data to the terminal by wireless communication. The storage unit 14 temporarily stores data received from the network.

FIG. 2 is a conceptual diagram of the control unit 11 in the present embodiment.
In the present invention, using the indicators (information) of the seven layers of the OSI (Open Systems Interconnection) reference model, the allocation order of each terminal is determined by an optimization method, and the optimized parameters are fed back to each layer. . Each layer performs transmission / reception using the optimum parameters fed back. In the present embodiment, optimization is performed using a multi-objective function based on the index of the application layer, the index of the transport layer, and the index of the physical layer. Also, the optimized parameters are fed back to the data link layer and the physical layer.

[Definition of metrics for each layer]
First, a method of defining indices for each layer used for optimization will be described.
(1) Definition of an index “QoS class” in the application layer.
First, the control unit 11 buffers the data to be transmitted to the terminal separately in the data link layer according to the attribute of the application used by the terminal. For example, the control unit 11 provides a priority buffer and a general buffer for each terminal in the data link layer. Then, when the application is an important application among delay-oriented applications or throughput-oriented applications, the control unit 11 buffers data to be transmitted to the terminal in a priority buffer. Examples of applications that place importance on delay include video streaming and VoIP (Voice over Internet Protocol). Examples of important applications that emphasize throughput include main objects such as e-mail and HTTP (Hypertext Transfer Protocol). The main object of http includes, for example, text (HTML sentence) written in HTML (HyperText Markup Language) on a Web page. On the other hand, when the application is an application that places importance on other throughput, the control unit 11 buffers data to be transmitted to the terminal in a general buffer. Another application that emphasizes throughput is, for example, an embedded object of http. Examples of the embedded object of http include an image referred to from an HTML sentence on a Web page. Note that the priority buffer and the general buffer include a retransmission buffer and a new buffer.

  The QoS class of the application of the user (terminal) i in the radio frame n is defined by Expression (1) based on the status of the priority buffer and the general buffer in the data link layer. The buffer delay is the number of delayed radio frames between the radio frame at the time of actual scheduling and the buffer reference radio frame of the data link layer. By using the QoS class as an index, scheduling according to the application used by the terminal becomes possible.

(2) Definition of transport layer index “segment average throughput”.
The control unit 11 acquires the segment throughput from each terminal in advance. A segment is a data unit of the transport layer (for example, a TCP packet, so-called maximum transmission unit MTU: Maximum Transmission Unit). For example, the segment throughput T _seg of the user i in the segment seg_id is calculated by Expression (2).

The segment average throughput A _seg of user i in the segment seg_id is defined by Equation (3) using T _seg . t _b is a forgetting factor. By using the segment average throughput as an index, scheduling can be performed according to the segment average throughput of each terminal.

(3) Definition of physical layer index “proportional fairness”.
The proportional fairness index M of the user i in the radio frame n is defined by Expression (4). T (i, n) is the average frequency utilization efficiency of user i in radio frame n. R (i, n-CIR delay) is the frequency utilization efficiency determined by the CIR of user i's radio frame (n-CIR delay). The CIR delay is the number of delayed radio frames between the radio frame at the time of actual scheduling and the CIR reference radio frame of the physical layer. The proportional fairness index enables scheduling to maintain fairness between terminals.

Further, the average frequency utilization efficiency T is updated by the equation (5). t _a is a forgetting counting.

[Determination of allocation order by optimization method]
Next, a procedure for determining the allocation order of each candidate user will be described. A candidate user is a user to whom radio resources are allocated. First, the control unit 11 normalizes each index. Next, the control unit 11 determines the allocation order of each candidate user based on the normalized index by an optimization method using a multi-objective function.

Hereinafter, normalization of each index will be described.
(1) Normalization of QoS class The QoS class is normalized to [0 to W _Class ]. Here, W _Class is a constant indicating the importance of the value f _Class obtained by normalizing the QoS class. The control unit 11 calculates f _Class of the candidate user i in the radio frame n by using the equation (6).

(2) Normalization of segment average throughput The segment average throughput is normalized to [0-W _T ]. Here, W _T is a constant indicating the importance of the value f _T obtained by normalizing the segment average throughput. A _MSID is a set of IDs for identifying candidate users. Control unit 11 calculates the _{f T} of the candidate user i in the radio frame n by Equation (7).

(3) Normalization of proportional fairness index The proportional fairness index is normalized to [0 to W _M ]. Here, W _M is a constant indicating the importance of the value f _M obtained by normalizing the proportional fairness index. Control unit 11 calculates the _{f M} of the candidate user i in the radio frame n by equation (8).

  Next, a method for determining an allocation order by an optimization method using a multi-objective function will be described. First, the control unit 11 calculates the value of the multipurpose function U for each candidate user. U of candidate user i in radio frame n is expressed by equation (9).

Next, the control unit 11 sets the candidate user allocation order in ascending order of the value of U. For candidate users having the same value of U, the order of candidate users having a high importance index is first. In the case of the following example, U (1, n) = 0.5 for candidate user 1 and U (2, n) = 0.5 for candidate user 2, and the value of U is the same. However, the value of the highest importance f _M, since towards the candidate user 2 is large, the allocation order candidate user 2 is first.
(Example)
W _Class = 0.5, W _M = 1, W _T = 0
(Candidate user 1) f _Class (1, n) = 0.5, f _M (1, n) = 0.5, f _seg (1, n) = 0
(Candidate user 2) f _Class (2, n) = 0, f _M (2, n) = 1, f _seg (2, n) = 0

[Parameter feedback to each layer]
Next, a method for feeding back parameters to each layer will be described.
(A) Feedback to Physical Layer The control unit 11 feeds back back (Backoff) back to the CIR used to determine the packet format in the physical layer, based on the QoS class of the assigned candidate user. The fed back Backoff is added to the CIR used to determine the packet format. When the QoS class is 1, the value of Backoff is Backoff _priority . On the other hand, when the QoS class is 0, the value of Backoff is Backoff _normal . That is, when the QoS class is 1, CIR _new = CIR _old + Backoff _priority . On the other hand, when the QoS class is 0, CIR _new = CIR _old + Backoff _normal . Backoff _priority and Backoff _normal are constants. In general, Backoff _priority ≧ Backoff _normal . CIR _new is the CIR after feedback. CIR _old is a CIR used to determine a packet format before feedback.

Further, the control unit 11 feeds back the parameter Gap to the maximum number of retransmissions of the physical layer based on the QoS class of the assigned candidate user. The feedback Gap is added to the maximum number of retransmissions. When the QoS class is 1, the value of Gap is Gap _priority . On the other hand, when the QoS class is 0, the value of Gap is Gap _normal . That is, when the QoS class is 1, the maximum number of retransmissions _new = the maximum number of retransmissions _old + Gap _priority . On the other hand, when the QoS class is 0, the maximum number of retransmissions _new = the maximum number of retransmissions _old + Gap _normal . Gap _priority and Gap _normal are constants. In general, Gap _priority ≧ Gap _normal . The maximum number of retransmissions _new is the maximum number of retransmissions after feedback. The maximum number of retransmissions “ _old” is the maximum number of retransmissions before feedback.

Further, the control unit 11 feeds back the parameter Diff to the physical layer retransmission delay for data transfer in the physical layer based on the QoS class of the assigned candidate user. The fed back Diff is added to the physical layer retransmission delay. When the QoS class is 1, the value of Diff is Diff _priority . On the other hand, when the QoS class is 0, the value of Diff is Diff _normal . That is, when the QoS class is 1, physical layer retransmission delay _new = physical layer retransmission delay _old + Diff _priority . On the other hand, when the QoS class is 0, physical layer retransmission delay _new = physical layer retransmission delay _old + Diff _normal . Diff _priority and Diff _normal are constants. Generally, Diff _priority ≦ Diff _normal . The physical layer retransmission delay _new is a physical layer retransmission delay after feedback. The physical layer retransmission delay _old is a physical layer retransmission delay before feedback.
Note that Backoff _priority , Backoff _normal , Gap _priority , Gap _normal , Gap _priority, and Gap _normal may be negative numbers.

(A) Feedback to data link layer The control unit 11 feeds back the parameter MACRT to the maximum number of retransmissions of the data link layer based on the QoS class of the assigned candidate user. The maximum number of retransmissions in the data link layer is the fed back MACRT. When the QoS class is 1, the value of MACRT is MACRT _priority . On the other hand, when the QoS class is 0, the value of MACRT is MACRT _normal . That is, when the QoS class is 1, the maximum number of retransmissions = MACRT _priority . On the other hand, when the QoS class is 0, the maximum number of retransmissions = MACRT _normal . MACRT _priority and MACRT _normal are positive constants. In general, MACRT _priority ≧ MACRT _normal .

FIG. 3 is a flowchart showing a flow of scheduling by the control unit 11 in the present embodiment.
First, in step S1, the control unit 11 updates the allocatable resource amount. That is, the control unit 11 calculates the amount of resources that can be allocated for data transmission, excluding physical layer retransmission resources and control message resources. In the next step S2, it is determined whether or not the updated allocatable resource amount is zero. If 0, the process is terminated. If it is not 0, the process proceeds to step S3.

  In the next step S3, the control unit 11 selects a user (terminal) as an allocation candidate. At this time, the control unit 11 sets a user whose accumulated amount in the priority buffer or general buffer in the data link layer is not 0 as a candidate user. The priority buffer or the general buffer holds data to be transmitted to the terminal. In the next step S4, the control unit 11 determines the candidate user assignment order by the optimization method using the multi-objective function described above.

In the next step S5, parameters are fed back to the physical layer and the data link layer using the QoS class of each candidate user. Specifically, when the QoS class is 0, the control unit 11 determines that CIR _new = CIR _old + Backoff _normal , maximum number of retransmissions _new = maximum number of retransmissions _old + Gap _normal, and physical layer retransmission delay _new = physical layer retransmission delay. and feeds back the _old + _{Diff normal} to the physical layer. Further, the control unit 11 feeds back the maximum number of retransmissions = MACRT _normal to the data link layer. On the other hand, when the QoS class is 1, the control unit 11 determines that CIR _new = CIR _old + Backoff _priority , maximum number of retransmissions _new = maximum number of retransmissions _old + Gap _priority, and physical layer retransmission delay _new = physical layer retransmission delay _old + Dif _priority. Is fed back to the physical layer. Further, the control unit 11 feeds back the maximum number of retransmissions = MACRT _priority to the data link layer.

  In the next step S6, the control unit 11 determines the resource amount to be allocated to each candidate user and ends the process. Specifically, the control unit 11 assigns necessary resource amounts for each candidate user in the order of assignment. The amount of resources required for each candidate user is calculated based on the amount of accumulated priority buffer and general buffer in the data link layer and the packet format determined by the CIR after feedback.

  As described above, according to the present embodiment, the user allocation order is determined by the optimization method using the multi-objective function using the index of each layer. In addition, the optimized parameters are fed back to each layer. Each layer performs transmission / reception using the optimum parameters fed back. As a result, radio resource allocation can be performed in consideration of higher layer information such as QoS required by the application layer. Furthermore, the parameters of each layer can be optimized.

  As described above, the embodiment of the present invention has been described in detail with reference to the drawings. However, the specific configuration is not limited to the above, and various design changes and the like can be made without departing from the scope of the present invention. It is possible to

For example, in this embodiment, the optimized parameters are fed back only to the physical layer and the data link layer, but may be fed back to other layers.
In the present embodiment, the indices of the application layer, transport layer, and data link layer are used. However, scheduling may be performed using indices of other layers.
In the present embodiment, the scheduling method described above is applied to the downlink of the OFDMA system, but it may be applied to other network systems.

It is a block diagram which shows the structure of the radio | wireless apparatus by one Embodiment of this invention. It is a conceptual diagram of the control part in this embodiment. It is a flowchart which shows the flow of the scheduling by the control part in this embodiment. It is the flowchart which showed the flow of the conventional scheduling.

Explanation of symbols

  DESCRIPTION OF SYMBOLS 1 ... Wireless apparatus 11 ... Control part 12 ... Wireless part 13 ... Network interface 14 ... Memory | storage part

Claims (14)

In a radio apparatus that allocates downlink radio resources to terminals,
A radio apparatus comprising: a control unit that determines an order of terminals to which radio resources are allocated based on an index for each terminal related to a communication protocol layer higher than a physical layer and a data link layer.   The radio apparatus according to claim 1, wherein the control unit normalizes each index according to importance.   The radio apparatus according to claim 1, wherein the index is a QoS class of an application layer or a segment average throughput of a transport layer.   The radio apparatus according to any one of claims 1 to 3, wherein the control unit feeds back an optimum parameter for each layer based on an index for each terminal.   The radio apparatus according to claim 4, wherein the control unit feeds back an optimum parameter for each layer based on a QoS class of an application layer.   The radio apparatus according to claim 5, wherein the control unit feeds back a parameter of CIR backoff, the maximum number of retransmissions, or a physical layer retransmission delay to the physical layer.   The radio apparatus according to claim 5, wherein the control unit feeds back a parameter of the maximum number of retransmissions to the data link layer. In a radio resource allocation method for allocating downlink radio resources to terminals,
A radio resource allocation method comprising: determining an order of terminals to which radio resources are allocated based on an index for each terminal related to a communication protocol layer higher than a physical layer and a data link layer.   The radio resource allocation method according to claim 8, wherein each index is normalized according to importance.   The radio resource allocation method according to claim 8 or 9, wherein the index is a QoS class of an application layer or a segment average throughput of a transport layer.   The radio resource allocation method according to any one of claims 8 to 10, wherein an optimum parameter for each layer is fed back for each terminal based on an index.   The radio resource allocation method according to claim 11, wherein an optimum parameter is fed back to each layer based on the QoS class of the application layer.   The radio resource allocation method according to claim 12, wherein CIR backoff, maximum number of retransmissions, or physical layer retransmission delay parameters are fed back to the physical layer.   The radio resource allocation method according to claim 12, wherein a parameter of the maximum number of retransmissions is fed back to the data link layer.

Images