KR101136715B1 - Packet Scheduling Algorithm considering a minimum delay time requirement for Non-realtime Traffic - Google Patents

Abstract

The present invention relates to a method of allocating resources in a wireless Internet environment such as a portable Internet, and more particularly, to a method of allocating resources in consideration of a minimum transmission delay time for providing an optimal non-realtime service to users. .

The method according to the invention comprises the steps of establishing a session between mobile terminals; Determining a priority value for each user according to adaptive scheduling using a maximum delay request time and an additional delay time for each user; And allocating frequency resources according to the determined priority value.

Resource allocation, non-real-time traffic, adaptive scheduling

Description

Resource Scheduling Algorithm Considering a Minimum Delay Time Requirement for Non-realtime Traffic

1 is a flowchart illustrating a frequency resource allocation method according to an embodiment of the present invention;

2 is a view for explaining a method for determining a resource allocation priority value according to an embodiment of the present invention;

3 is a view showing a comparison result of the yield of the resource allocation method according to an embodiment of the present invention and the resource allocation method according to the prior art,

4 is a view showing a comparison result of the delay loss probability of the resource allocation method according to an embodiment of the present invention and the resource allocation method according to the prior art;

5 is a view showing another example of a comparison result of the delay loss probability of the resource allocation method according to an embodiment of the present invention and the resource allocation method according to the prior art;

6 is a view showing another example of a comparison result of the delay loss probability of the resource allocation method according to an embodiment of the present invention and the resource allocation method according to the prior art.

The present invention relates to a resource allocation method in a wireless Internet environment, such as a mobile Internet, and more particularly, to a resource allocation method considering a minimum transmission delay time for providing an optimal non-realtime service to users. It is about.

Recently, various resource allocation methods have been studied to perform resource allocation while satisfying fairness among users in mobile communication systems, and the necessity for satisfying service quality of non-real-time service users is increasing.

Conventional wireless packet scheduling algorithms considering non real-time services are usually designed to increase system capacity or to ensure fairness among users. Unlike real-time services, "A. Jalali, R. Padovani, and R. Pankaj," Data Throughput of CDMA-HDR a High Efficiency-High Data Rate Personal Communicaiton Wireless System, "Proceeding of IEEE VTC 2000, pp. 1854-1858, vol. 3, May 2000. The PF method is mainly used. Non-real-time including the above-described PF method Existing wireless packet algorithms considering services are able to allocate resources fairly, but do not consider transmission delays, so some users may be delayed due to their priority, resulting in poor service satisfaction.

In real-time services, the transmission delay is strictly limited. There is also a requirement for delay for non-real-time traffic, which requires the development of a scheduling algorithm that can flexibly adjust the tradeoff between delay performance and system yield.

Accordingly, an object of the present invention is to provide a method for allocating a resource suitable for a non-real time service.

 An object of the present invention is to provide an optimal resource allocation method for a non-real-time traffic service by applying an adaptive scheduling method for improving system capacity while providing a soft QoS of a form suitable for a non-real-time service.

The method according to the present invention for achieving the object of the present invention comprises the steps of establishing a session between the mobile terminal; Determining a priority value for each user according to adaptive scheduling using a maximum delay request time and an additional delay time for each user; And allocating frequency resources according to the determined priority value.

delete

 The above objects, features and advantages will become more apparent from the following detailed description taken in conjunction with the accompanying drawings. First of all, in adding reference numerals to the components of each drawing, it should be noted that the same components have the same number as much as possible even if displayed on different drawings. In addition, in describing the present invention, when it is determined that the detailed description of the related known technology may unnecessarily obscure the subject matter of the present invention, the detailed description thereof will be omitted. Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings.

Real-time traffic services strictly require transmission delays, which reduces the possibility of transmission delays. However, the non-real-time traffic service does not require a strict delay performance, so that loss can be tolerated as needed. In this case, a gain-and-loss relationship between loss performance and system yield according to a target delay time can be considered in scheduling. Therefore, the principle of the present invention is to increase the yield by preferentially selecting a user with a good channel condition while delaying the service of the user, even if a loss occurs when there is a user with a bad channel condition among the users near the target delay time. It is to let.

1 is a flowchart illustrating a frequency resource allocation method according to an embodiment of the present invention.

Referring to FIG. 1, frequency resource allocation is performed by a base station of a portable Internet system having a processor in charge of resource allocation. First, in step 110, a session is established between the base station and the terminal. In step 120, the base station determines the transmission rate determined by the maximum delay request time, channel status for each user, delay loss probability, and channel status. Thereafter, in step 130, the base station determines a priority value by an adaptive scheduling method described later. Finally, in step 140, the base station allocates frequency resources according to the determined priority value. Hereinafter, the adaptive scheduling method of step 130 will be described in detail.

<Adaptive Exponential Scheduling>

를 기준으로 Max C/I에 의한 자원할당을 적응적으로 가능하도록 MDS 스케줄링을 변형한 Adaptive Exponential 스케줄링을 제안한다.

는 이동통신 시스템이 트래픽 성질에 따라 임의로 결정한다.In the embodiment of the present invention, considering the characteristics of the non-real-time traffic service described above,

Based on this, we propose Adaptive Exponential Scheduling with modified MDS Scheduling to adaptively allocate resources by Max C / I.

Is arbitrarily determined by the mobile communication system according to the traffic characteristics.

2 is a diagram illustrating a method for determining a resource allocation priority value according to an embodiment of the present invention.

) 에 사용자 별로 추가적인 지연 시간

를 허용함으로써, 우선권을 부여하는 방식에 있어서 보다 유연한 대처가 가능하다. As shown in FIG. 2, the Adaptive Exponential Scheduling method uses the maximum delay request time required by the service.

Additional latency per user

This allows more flexibility in the way of giving priority.

동안 채널 상태가 우수한 사용자에 대해 우선권을 양보함으로써 시스템의 수율을 증가시킨다. 이때, 우선권을 제어하기 위해

값은 하기 <수학식 1> 같이 설정한다.That is, referring to FIG. 2, the priority curve is shifted to the right for a user whose channel condition is poor while allowing some delay loss.

Increase yield of the system by yielding priority to users with good channel condition. At this time, to control the priority

The value is set as in <Equation 1>.

는 시간 t에서 1bit를 전송할 때 지연 손실이 발생할 확률이고,

는 시간 t에서 채널의 상태에 의해 결정된 전송률을 나타낸다. 한편, 알고리즘이 시스템(기지국)에서 운용될 때

값은 전체 전송된 비트 수와 지연 손실이 발생한 비트수의 비율로써 매 프레임 마다 갱신된다. In Equation 1, α is a system control parameter,

Is the probability of delay loss when transmitting 1 bit at time t,

Denotes the transmission rate determined by the state of the channel at time t. On the other hand, when the algorithm is operated in the system (base station)

The value is updated every frame as the ratio of the total number of transmitted bits to the number of bits where the delay loss occurred.

값을 적절하게 설정하고 매 프레임 마다 갱신함으로써 간접적으로 우선권을 부여하게 된다. 즉,

값이 크고, 손실이 발생하는 비율

가 큰 사용자에 가장 우선적으로 자원을 할당하고,

값이 작고 손실이 발생하는 비율

이 낮은 사용자는 가장 낮은 자원할당 우선순위를 가지게 된다. 이것은 도 2에 도시된 바와 같이,

값이 작으면 우선권 커브가 좌측으로 이동하게 되고, 그 반대이면 우선권 커브가 우측으로 이동하게 되기 때문이다. 또한, 임의의 사용자의 우선권 커브가 우측으로 이동했다 하더라도 채널 상태가 우수하면 우선적으로 서비스 받을 수 있음을 알 수 있다.

값은 수율과 지연 손실간의 득실을 최적화하기 위한 수단으로 사용되며, 이를 통해 비실시간 트래픽의 유연한 서비스 요구 조건에 대한 특성을 반영할 수 있다. Adaptive Exponential scheduling 방식을 의사코드로 기술하면 하기 <표 1>과 같다.According to the situation as shown in Equation 1 above

Priority is given indirectly by setting the value appropriately and updating every frame. In other words,

High value, loss rate

Assign resources first to large users,

Small value and loss rate

This lower user will have the lowest resource allocation priority. This is shown in Figure 2,

If the value is small, the priority curve is shifted to the left, and vice versa, the priority curve is shifted to the right. In addition, even if the priority curve of any user has moved to the right, it can be seen that the service can be preferentially provided that the channel condition is excellent.

The value is used as a means of optimizing the tradeoff between yield and delay loss, which reflects the characteristics of flexible service requirements for non-real-time traffic. The adaptive exponential scheduling method is described in pseudo code as shown in Table 1 below.

<Experiment Result for Example of the Present Invention>

3 is a view showing a comparison of the yield of the resource allocation method according to an embodiment of the present invention and the resource allocation method according to the prior art, Figures 4 to 6 is a resource allocation method according to an embodiment of the present invention and the prior art A diagram showing a comparison result of the delay loss probability of the resource allocation method according to the technique. Hereinafter, the experimental results will be described in detail with reference to FIGS. 3 to 6.

In the present invention, the experiment for comparing with the performance of the prior art simultaneously considered the uplink / downlink based on the portable Internet system standards. The network consists of 19 cells, and each cell uses a 3-sector antenna to increase frequency efficiency. In order to make it similar to the real channel environment, the path loss model based on the distance from the base station, the shadowing model based on the obstacle and the moving speed, and the multipath loss model are reflected.

And, all the user's moving speed was set to 60km / h environment. The channel used was the performance of the network average yield per sector using only the diversity subchannel (Normal Subchannel). In the present invention, the performance of the scheduling algorithm is measured through system level simulation of the ultra-high speed portable Internet system (Korea Information Technology Association "2.3GHz portable Internet standard media access control layer" Information and Communications Organization Standard PG302, Jun, 7, 2004.). Channel models and system parameters were applied as shown in Table 2 below.

In the simulation, the system calculates and reflects the user's required CINR value in every frame according to the AMC mode, and assumes the total power of the base station is 20 watts. The channel model considers both the COST-231 path loss model, the shadowing model, and the multipath channel model. In addition, the distribution deviation according to the shadow fading is set to 8dB, the correlation distance is set to 50m. In addition, the 3-sector cell was considered, and all users were generated uniformly in the cell, and these users were allowed to move at a speed of 60 km / h. Meanwhile, hard handover is applied to a user moving to another cell. In addition, it is assumed that a transmission error on a wireless channel does not occur, and thus no separate ARQ protocol is applied. Therefore, an ideal channel without retransmission overhead according to the ARQ protocol is considered. Therefore, this simulation result corresponds to the highest possible performance. The simulation time is shown by measuring the performance of 5000 frames, that is, the time progressed for 25sec. Meanwhile, various variable values used in the computational experiments for the present invention are as described in Table 3 below. AMC (Adaptation of Modulation & Coding) can be supported by the CINR value according to the channel state of the user in the physical layer. Therefore, in this simulation, the transmission mode determination based on the CINR value of the user is determined by Table 3 below, and transmission and reception with the base station is performed in the corresponding AMC mode.

The traffic used in the comparative computation experiment of the resource allocation method and the prior art according to an embodiment of the present invention used Ethernet traffic. We analyzed the performance by increasing the Ethernet traffic. The arrival time between successive packets of the used traffic follows the Pareto distribution generally considered in Internet traffic. This is because the Pareto distribution is a good representation of the traffic's self-similarity and burst characteristics. The length of the packet is based on the measured distribution in Ethernet traffic (R. Gusella, "A Measurement Study of Diskless Workstation Traffic on an Ethernet," IEEE Transactions on Comm., Pp. 1557-1568, vol. 38, no. 9, September 1990.) (W.-M. Yin and Y.-D. Lin. "Statistically optimized Minislot Allocation for Initial and Collision Resolution in Hybrid Fiber Coaxial Networks," IEEE Journal on Selected Areas in Communication, pp 1764-1773, vol. 18, no. 9, September 2000. The Pareto distribution is determined by shape parameter x and location parameter y, and the probability distribution function according to Equation 2 is as follows.

Here, the average arrival interval is expressed by Equation 3 below.

Here, λ is the packet arrival rate. In this simulation, the packet arrival rate λ is set to 0.075 packets / slot, and λ is 0.036 and 0.1 packets / Compare with slots. In addition, x = 1.3 was used.

On the other hand, the distribution of the average length of the packet was used to obtain the statistics obtained by analyzing the traffic generated in Ethernet, as shown in Table 4 below.

Traffic generated through the above-described process is converted into the size of the MAC PDU defined in the system, and padding if necessary to fit the fixed length packet in the radio section.

3 to 6, the embodiment and the comparison object of the present invention is a propagation method of PF (Proportion Fairness) resource, M-LWDF (M. Andrews, K. Kumaran, K. Ramanan, A. Stolyar, P. Whiting, and R. Vijayakumar, "Providing Quality of Service over a shared wireless link," IEEE Communications Magazine, pp. 150-154, February 2001.) and MDS (Simon Shin, Byeung-Cheol Kim, Jae-hwang) Yu, and Dongwoo Kim, "Analysis of 1xEV-DO Packet Scheduling Algorithms Considering QoS," Proceedings of CIIT 2003, 2003.

가 0.6sec, 1.2sec, 1.8sec 일 때 각각 6.86Mbps, 7.54Mbps, 8.05Mbps로서 수율이 증가함을 보인다. 본 발명의 실시예에 따른 자원 할당 방식을 적용하지 않은 경우에는 PF 방식과 M-LWDF 방식이 각각 5.91Mbps와 5.7Mbps의 수율이 나왔다. 따라서, 본 발명을 적용함으로써 최대 33%의 수율 상승을 가져왔다. Referring to FIG. 3, the yield in the case of applying the resource allocation method according to the embodiment of the present invention is

The yield is increased to 6.86Mbps, 7.54Mbps and 8.05Mbps, respectively at 0.6sec, 1.2sec and 1.8sec. When the resource allocation scheme according to the embodiment of the present invention is not applied, the PF scheme and the M-LWDF scheme yield 5.91 Mbps and 5.7 Mbps, respectively. Thus, application of the present invention resulted in a yield increase of up to 33%.

를 길게 할수록 Max C/I 방식이 적용되는 구간이 길어지도록 설계되었기 때문이다.The reason for this higher yield is that the proposed scheduling scheme

This is because the longer the is, the longer the section to which the Max C / I method is applied.

= 0.6sec 일때 본 발명의 실시예에 따른 자원 할당 방식을 적용한 경우와 적용하지 않은 경우에 대해 비교한 것이다. 본 발명의 실시예에 따른 자원 할당 방식을 적용하지 않은 경우에는 지연 손실 확률이 10% 이하가 되도록 만족시키려면 시스템에서 수용 가능한 수가 20명 이하가 된다. 본 발명의 실시예에 따른 자원 할당 방식을 적용한 경우에는 지연 손실 확률이 10% 이하가 되도록 만족시키려면 시스템에서 24명까지 수용 가능한 것을 알 수 있다. 따라서, 본 발명을 적용하면 시스템에서 20% 이상의 사용자 수를 수용할 수 있다는 것을 알 수 있다.Referring to Figure 4,

= 0.6sec, a comparison is made between the case where the resource allocation method according to the embodiment of the present invention is applied and the case where the resource allocation method is not applied. If the resource allocation method according to the embodiment of the present invention is not applied, the number acceptable to the system is 20 or less to satisfy the delay loss probability of 10% or less. When the resource allocation method according to the embodiment of the present invention is applied, it can be seen that up to 24 people can be accommodated in the system to satisfy the delay loss probability of 10% or less. Thus, it can be seen that applying the present invention can accommodate 20% or more users in the system.

= 1.2sec 인 경우이고, 도 6을 참조하면,

= 1.8sec일 때, 지연 손실 확률의 비교도이다. 본 발명의 실시예에 따른 자원 할당 방식은

가 길수록 다른 방식보다 현저히 지연에 대한 손실이 감소함을 알 수 있다. 반면에,

가 증가함에 따라 PF 방식에 대한 손실 확률은 거의 변하지 않는다. 이것은 PF 방식이 시스템 수율에서는 adaptive exponential 스케줄링 방식과 유사한 성능을 나타내지만, 사실상 지연 성능에 따른 우선권 부여를 고려하고 있지 않기 때문이다.5,

= 1.2sec, and referring to FIG. 6,

= 1.8 sec, the comparison of the delay loss probability. Resource allocation scheme according to an embodiment of the present invention

It can be seen that the longer the loss of the delay is significantly lower than other methods. On the other hand,

As is increased, the loss probability for the PF method hardly changes. This is because the PF scheme shows similar performance to the adaptive exponential scheduling scheme in system yield, but it does not consider prioritization according to delay performance.

While the invention has been described above based on the preferred embodiments thereof, these embodiments are intended to illustrate rather than limit the invention. It will be apparent to those skilled in the art that various changes, modifications, or adjustments to the above embodiments can be made without departing from the spirit of the invention. Therefore, the protection scope of the present invention should be construed to include not only the appended claims but also all of the above changes, modifications or adjustments.

As described above, in the present invention, when the system requires the satisfaction of the maximum transmission delay request time with respect to the transmission delay for the non-real-time service, the number of users that satisfies the maximum can be accommodated more than the conventional scheme. Therefore, there is an effect of increasing the capacity of the network.

Claims (5)

In the resource allocation method for the non-real-time traffic service in the base station of the mobile communication system, Establishing a session between mobile terminals; Determining a priority value for each user according to adaptive scheduling using an additional delay time and a maximum delay time required for each user according to a channel state; And Allocating a frequency resource according to the determined priority value; Resource allocation method for a non-real-time traffic service comprising a. The method of claim 1, The additional delay given to each user of the adaptive scheduling is A method for allocating a resource for a non-real-time traffic service, characterized in that it is determined by a predetermined system adjustment parameter, a probability of delay loss in an arbitrary time interval, and a transmission rate determined according to a channel state in an arbitrary time interval. The method of claim 1, The additional delay given to each user of the adaptive scheduling is Equation

, Where α is a system control parameter,

Is the probability of delay loss when transmitting 1 bit at time t,

Is the rate determined by the state of the channel at time t, The resource allocation method for a non-real time traffic service, characterized in that determined by. The method according to any one of claims 2 to 3, wherein a high priority value is given when a transmission rate determined according to a channel state in a certain time interval is large and a delay loss probability in a certain time interval is large. A resource allocation method for a non-real time traffic service. delete

Images