基于软件定义无线网络的云业务上行调度方案

doi:10.11959/j.issn.1000-436x.2016232

摘要/Abstract

摘要：

针对无线云计算的业务特点，基于软件定义无线网络(SDWN, software-defined wireless network)，提出了一种新的能适配云业务特点的LTE上行调度方案，该方案首先将资源分配问题归纳成二进制整数规划问题进行求解，然后利用动态规划计算出每个调度时隙内云业务的最佳发送速率，最后通过SDWN框架下的QoS控制方法动态调节云业务的发送速率，使云业务的发送速率与当前信道条件成正比，从而在保障多业务速率需求的同时，满足云业务能量消耗最小的需求，通过仿真验证算法的性能。

关键词: SDWN, 无线云, LTE调度, 最优化

Abstract:

A LTE uplink scheduling scheme matching the features of wireless cloud services was proposed for the SDWN (software-defined wireless network). The scheme first solved the resource allocation problem by using the binary integer programming method, and then calculated the optimal transmission rate of cloud services in each time slot using the method of dynamic programming, finally adjusted the transmission rate of cloud services proportional to the current channel status using QoS control method in the framework of SDWN. The proposed scheme minimizes the energy con-sumption of cloud services while ensuring the transmission rate demand of multiple services. The performance of the al-gorithm is verified by simulation.

Key words: SDWN, wireless cloud, LTE scheduling, optimization

潘甦,肖榜,陈宇青,刘胜美. 基于软件定义无线网络的云业务上行调度方案[J]. 通信学报, 2016, 37(12): 21-31.

Su PAN,Bang XIAO,Yu-qing CHEN,Sheng-mei LIU. Uplink scheduling scheme for cloud service based on software-defined wireless network[J]. Journal on Communications, 2016, 37(12): 21-31.

图/表 11

表1

算法中主要符号定义与描述"

符号	定义
$R_{i}^{0}$	云业务时隙t的速率需求
$R_{m}^{0}$	用户m的速率需求
KA	所有满足连续性限制的分配方案集合
m A	用户m的子信道分配矩阵
C	可能的分配方案数
r	速率矩阵
r_m,j	用户m在采用A_m中第j列对应的分配方案时获得的速率
K_m,j	A _m中第j列对应的分配方案中分配给用户m的子信道的个数
e	目标指数矩阵
_m,j e	用户m在采用A_m中第j列对应的分配方案时的目标指数
Χ	信道分配方案矩阵
_m,j x	用户m在采用A_m中第j列对应的分配方案时信道分配情况
$E r_{t} (b (t), \bar{h_{i, K_{i}}} (t))$	云业务时隙t的能量消耗
$Y_{t} (b (t), \bar{h_{i, K_{i}}} (t))$	云用户在时隙t到1的最小能量消耗和
${\bar{Y}}_{t} (b (t), \bar{h_{i, K_{i}}} (t))$	云用户在时隙t到1的最小平均能量消耗和
x v	x时隙内信道状态的估计
G(v _t ,?,v ₁ )	x v的几何平均值，对信道状态的估计

表1

图1

图2

表2

图3

图4

图5

图6

图7

图8

表3

参考文献 18

[1]	3GPP TS 23.203. Policy and charging control architecture[S]. 2013.
[2]	ABU-ALI N , TAHA A E M , SALAH M . Uplink scheduling in LTE and LTE-advanced: tutorial, survey and evaluation framework[J]. IEEE Communications Surveys ＆ Tutorials, 2014,16(3):1239-1265.
[3]	LEE J , JINDAL N . Energy-efficient scheduling of delay constrained traffic over fading channels[J]. IEEE Transactions on Wireless Com-munications, 2009,8(4):1866-1875.
[4]	ZHANG W , WEN Y , GUAN K . Energy-optimal mobile cloud com-puting under stochastic wireless channel[J]. IEEE Transactions on Wireless Communications, 2013,12(9):4569-4581.
[5]	NUNES B A A , MARC M . A survey of software-defined networking:past, present and future of programmable networks[J]. IEEE Commu-nications Surveys ＆ Tutorials, 2014,16(3):1617-1634.
[6]	PENTIKOUSIS K , WANG Y , HU W . Mobileflow: toward soft-ware-defined mobile networks[J]. IEEE Communications Magazine, 2013,51(7):44-53.
[7]	GUERZONI R , TRIVISONNO R , SOLDANI D . SDN-based archi-tecture and procedures for 5G networks[C]// IEEE 5GU'14. 2014:209-214.
[8]	AISSIOUI A , KSENTINI A , GUEROUI A M . Toward elastic distrib-uted SDN/NFV controller for 5G mobile cloud management sys-tems[J]. IEEE Access, 2015,3:2055-2064.
[9]	RIGGIO R , MARINA M K , RASHEED T . Interference management in software-defined mobile networks[C]// IEEE IM'15. 2015:626-632.
[10]	ZHONG L , NAKAUCHIK , SHOJI Y . Performance analysis of appli-cation-based QoS control in software-defined wireless net-works[C]// IEEE IWCMC'14. 2014:464-469.
[11]	KIM J , KIM D , HAN Y . Proportional fair scheduling algorithm for SC-FDMA in LTE uplink[C]// IEEE GLOBECOM'12. 2012:4816-4820.
[12]	YANG H , REN F , LIN C . Frequency-domain packet scheduling for 3GPP LTE uplink[C]// IEEE INFOCOM'10. 2010:1-9.
[13]	DELGADO O , JAUMARD B . Scheduling and resource allocation for multiclass services in LTE uplink systems[C]// IEEE WiMob'10. 2010:355-360.
[14]	WEN P , YOU M , WU S . Scheduling for streaming application over wideband cellular network in mixed service scenarios[C]// IEEE MILCOM'07. 2007:1-7.
[15]	SOKMEN F I , GIRICI T . Uplink resource allocation algorithms for single-carrier FDMA systems[C]// IEEE EW' 10. 2010:339-345.
[16]	MYUNG H G , LIM J , GOODMAN D J . Single carrier FDMA for uplink wireless transmission[J]. IEEE Vehicular Technology Magazine, 2006,1(3):30-38.
[17]	LEE K D , LEUNG V . Capacity planning for group-mobility users in OFDMA wireless networks[J]. EURASIP Journal on Wireless Com-munications and Networking, 2006(2):59-59.
[18]	WONG I C , OTERI O , MCCOY W . Optimal resource allocation in uplink SC-FDMA systems[J]. IEEE Transactions on Wireless Com-munications, 2009,8(5):2161-2165.

参数	值
基站范围半径	800m
终端与基站之间最小距离	90m
小区内用户分布	均匀分布
调度周期	1ms
路径损耗	(128.1+37.6 log D)dB
阴影效应	对数正态分布(μ=0, σ=8)
多径衰落	瑞利衰落(μ=0, σ=1)
终端最大发送功率Pu	200mW
噪声功率谱密度N 0	161dBm/Hz
子信道带宽W_s	180kHz
子信道数K	20
用户数M	15
一般用户速率要求0mR	64kbit/s,256kbit/s

比较方面	现有相关工作	本文工作
应用场景	传统网络框架下的LTE系统	SDWN网络框架下的LTE系统
优化目标	单一目标（如最大化吞吐量、最小化功率等）	多目标（如最大化吞吐量、最小化云业务上传耗能等）
限制条件	难以考虑差异较大的不同QoS要求	可以满足不同业务的QoS要求
优化方法	对系统内所有业务采用相同的优化方法（如贪婪算法、注水算法等）	采用分步优化的方法，首先利用0-1整数规划算法优化系统吞吐量，然后利用动态规划算法优化云业务耗能
方案效果	能达到优化目标要求，但不能优化云业务耗能	能够有效减少云业务耗能
适配新业务的能力	差	好