2. 西安电子科技大学 综合业务网理论及关键技术国家重点实验室, 西安 710071
在保证主用户服务质量的前提下,将接入点(AP)以认知用户的身份接入到主用户频谱,合理地分享主用户信道资源.将信道分配问题建模为整数线性规划问题,提出一个混合最优信道分配算法(HOCA),该算法基于一种有权连接图的最小生成树,可以快速准确地在主用户和认知AP频谱中为主用户和认知AP分配干扰最小的信道.最后,将HOCA算法和其他2种信道分配算法,次最优机会信道分配和次最优信道分配-2算法进行了数值分析,验证了HOCA在吞吐量性能的有效性.
2. State Key Laboratory of Integrated Services Networks, Xidian University, Xi'an 710071, China
A solution is put forward that makes access point (AP) access into the primary user's channel in the identity of cognitive user under the precondition of guaranteeing primary user's quality of service, so as to share the channel resource of primary user effectively. It models the channel assignment as an integer linear programming issue and puts forward a hybrid optimal channel assignment algorithm (HOCA). Based on minimum spanning tree algorithm of connected weighted graph, the algorithm can allocate the minimum interfered channel for primary user and cognitive AP fast and efficiently in its own frequency band of primary user and cognitive AP. HOCA algorithm and other two channel assignment algorithms, near-optimal opportunistic channel allocation and near optimal frequency assignment-2 are analyzed. It is testified that HOCA algorithm can effectively improve the throughput of cognitive user.
目前,无线信道干扰问题已经成为影响高密度多接入点共存下吞吐量的一个热点问题[1-9].解决信道干扰问题主要有2种办法:1) 在热点小区内由网络管理人员对免费频段信道进行统一规划分配,尽量让相邻的AP工作在正交信道,避免同频信道干扰,但是该方法实现起来不确定因素非常多.首先,非授权频段资源非常有限;其次,由于无线AP分别归属不同的所有者,网络管理人员不太可能与这些所有者进行协商,并且无线AP接入是动态的,所以该方案实现起来难度较大. 2) 通过认知无线电方式接入授权频段拓展当前无线局域网(WLAN,wireless local area network)的工作频段,这种方式目前得到越来越多人的重视.
综上,提出一种WLAN AP以认知无线电方式接入主用户授权频段进行合理信道分配的方法,该方法在保证主用户网络服务质量(QoS,quality of service)的前提下,根据当前工业科技医疗(ISM,industrial scientific medical)频段的资源竞争以及干扰情况,可以选择以机会方式接入主用户频谱资源,扩大AP工作的信道范围,减小信道分配干扰,从而最大化认知用户吞吐量.
1 混合网络建模图 1所示为主用户网络和认知用户网络共存时的拓扑结构.其中,主用户网络由主用户基站和主用户客户端组成,认知用户网络由认知AP和认知节点组成.主用户网络工作在授权频段,而认知用户网络一般工作在非授权频段,这里特指ISM(industrial scientific medical)频段.当ISM频段频谱资源紧张并且主用户频段有空闲频谱时,认知AP采用机会方式接入主用户网络信道,但是一旦主用户需要授权频谱资源时,认知AP应无条件地让出主用户信道.
认知AP一般工作在ISM频段,主用户工作在授权频段.假设ISM频段FISM有11个信道{1, 2, …, 11},1、6和11为正交信道,而主用户频段FP有k个信道{M1, M2, …, Mk},信道带宽H MHz,相邻正交信道间隔数为L.
本研究采用无向权重连接图G{V, E}来表示混合网络拓扑,v∈V表示需要分配信道节点,具体为主用户基站和认知AP,e∈E表示节点之间的连线,且所带权重跟连接的2个节点的距离成正比.
如果节点v分配了信道f,用f∈F(v)表示,那么对于2个已经分别分配到信道fm和fn的节点m和n,它们之间的互干扰水平通过干扰惩罚Gm, n(fm, fn)来确定.
(1) |
(2) |
其中:Amn表示第n个AP的干扰区域和第m个AP的通信区域重叠的区域;AC, m表示第m个AP的通信区域;ρ(fm, fn)表示信道重叠因子;O是正交信道间隔数,对ISM频段而言,O=5,对主用户频段而言,O=L.
采用二进制整数线性规划来描述信道分配问题.
使用二进制变量表示某一信道是否分配给某一节点(主用户基站和认知AP),如信道f分配给第v个节点,可以定义为
(3) |
每个用户应分配且仅能分配1个信道,因此用户的信道分配条件为
(4) |
其中:F表示ISM频段和主用户频段可用信道数之和,N为认知AP和主用户基站数目之和.
对于两节点v和w互干扰惩罚Gv, w,存在一定的干扰惩罚上限Gmax,若Gv, w>Gmax,那么这一对节点的信道分配是不合理的,需要进行重新分配,信道分配限制条件为
如果
(5) |
设定AP工作在ISM频段时,满足AP平均吞吐量的最大互干扰惩罚为Gc_max,在主用户QoS存在前提下,影响主用户和认知用户性能的混合网络最大互干扰惩罚为Gp_max .
2.2 混合最优信道分配算法AP和主用户基站为图G{V, E}的节点,存在重叠域的AP之间的连线为边,AP的距离为边的权重.构造集合Vnew和Enew,分别表示已分配信道的节点和已调用的边. HOCA信道分配伪代码如下.
1 Vnew={x}, where x is the base station, fx=31
2 Vnew={x, y}, where y is the node that is the nea-
rest AP from the base station, if there are many
APs with the same distance, select one randomly,
fy=1
3 while V≠Φ do
4 Choose edge (u, v) from E with minimal
weight, that u in V has the most overlapping area
with v in V new , E new =(u, v)
5 case1 v has only one neighbor u
6 if u’ s channel in primary band then
7 fv=1
8 else
9 fv=(fu+5) mod11
10 end if (6)
11 case2 v has only two neighbors u, w
12 if at least one neighbor’s channel in primary
band then
13 Jump to case 1
14 else
15 Choose fv s.t. min (|fv-fu|, |fv-fw|)=5
16 end if (12)
17 case 3 v has three or more (choose the three
with minimal weight) neighbors u, w, y
18 if at least one neighbor’s channel in primary
band then
19 Jump to case 2
20 else
21 Choose fv s.t. min (|fv-fu|, |fv-fw|,
|fv-fy|)=g, firstly g=5
22 if fv has solution then
23 Choose the channel (has multiple so-
lutions choose the one that cause
minimizes IP(interference penalty))
24 else
25 g=g-1 until g=1, jump to line 21
26 end if (22)
27 end if (18)
28 if all pair of nodes’IP is less than its threshold
then
29 Assign channel as calculated in ISM band
30 else
31 if all primary channel are unavailable then
32 Assign channel as calculated in ISM
band
33 else
34 Assign channel as line4-23 in primary
band
35 if all pair of nodes’ IP is less than its
threshold then
36 Assign channel as calculated in pri-
mary band
37 else
38 Assign channel as calculated in ISM
band
39 end if (35)
40 end if (31)
41 end if (28)
42 Move v from V to V new , move (u, v) from E to E new
43 end while (3)
3 数值分析为了更好地对所提算法进行性能分析,将HOCA算法和文献[5]中的次最优机会信道分配(NOOCA,near-optimal opportunistic channel allocation)以及文献[4]中的次最优信道分配-2(NOFA-2,near optimal frequency assignment-2) (ISM)算法进行数值分析.结果如图 2所示.
图 2中3种算法在节点互干扰未达到Gc_max之前,平均干扰趋势相同,这是由于在AP数目比较少的情况下,ISM频段并没有出现资源紧张.特别地,当不多于3个的AP进行信道分配时,系统的平均干扰为0,这是因为ISM频段有3个正交信道,分别是1、6和11,AP使用正交信道时,相互之间不存在干扰.
当超过Gc_max之后,随着AP数目增多,NOFA-2算法的平均干扰一直比NOOCA和HOCA算法要高.这是因为NOFA-2算法只针对ISM频段进行信道分配,而在ISM频段的AP干扰已经超过了满足平均吞吐量规定的Gc_max,在此基础上再继续增加AP,只能增加信道干扰,造成平均干扰的增加.而NOOCA和HOCA算法的平均干扰相对较小,这是因为这2种算法都考虑到了主用户频段,特别地,HOCA算法在超过Gc_max之后新增2个AP,平均干扰不变,这是因为主用户正交信道数有3个,分别是信道31、36和41,31号正交信道初始分配时已经分配给基站,另外2个按照算法分配给认知AP,而NOOCA算法是选择受到重叠区域最严重(干扰超过ISM频段阈值)的3个AP分配主用户正交信道,所以即使新增3个AP,总的平均干扰不变. HOCA算法在不影响主用户QoS的前提下,即不超过Gp_max之前,在主用户频段最大可允许接入的AP数为15个.在主用户正交信道分配完毕之后,HOCA算法继续在主用户和ISM频段中进行干扰最小原则分配,NOOCA算法则只在主用户频段中进行信道分配,在未超过Gp_max之前,NOOCA算法整体干扰水平比HOCA算法要略高.
当NOOCA和HOCA算法超过Gp_max之后,HOCA为了保护主用户QoS,在主用户频段不再增加AP信道的分配,而是在ISM频段进行AP分配,而NOOCA由于没有考虑主用户QoS的问题,新增加的AP还是继续在主用户频段进行信道分配,所以在这个区域HOCA算法的平均干扰比NOOCA算法略高.
4 结束语提供了一种拓展ISM频谱资源的思路,在不影响主用户QoS的前提下,WLAN AP以认知无线电的方式机会接入主用户频谱,让AP可以在ISM和主用户可用频谱中进行信道分配.将混合网络信道分配问题定义为整数线性规划问题,并通过基于互干扰最小限制的最小生成树方法分配主用户和认知AP信道,兼顾了主用户QoS并提高了认知用户的吞吐量,较好地提高了频谱资源利用率.
[1] | Zhang Heli, Ji Hong, Ge Wendong. Channel assignment with fairness for multi-AP WLAN based on distributed coordination function[C]//2011 IEEE Wireless Communications and Networking Conference. Cancun, Mexico: IEEE Press, 2011: 392-397. http://dblp.uni-trier.de/db/conf/wcnc/wcnc2011 |
[2] | Abbasi S, Kalhoro Q, Kalhoro M A. Efficient use of partially overlapped channels in 2.4 GHz WLAN backhaul links[C]//2011 International Conference on Innovations in Information Technology. Abu Dhabi, United Arab Emirates: IEEE, 2011: 7-11. |
[3] | Chan E C L, Baciu G, Mak S C. Effect of channel interference on indoor wireless local area network positioning[C]//2010 IEEE 6th International Conference on Wireless and Mobile Computing, Networking and Communications. Niagara Falls, Canada: IEEE, 2010: 239-245. |
[4] | El-Hajj W, Alazemi H. Optimal frequency assignment for IEEE 802.11 wireless networks[J]. Wireless Communications and Mobile Computing, 2009, 9(1): 131–141. doi: 10.1002/wcm.v9:1 |
[5] | Novillo F, Churchman M, Ferrús R, et al. A channel allocation algorithm for OSA-enabled IEEE 802.11 WLANs[C]//Proceedings of the 2009 6th International Symposium on Wireless Communication Systems. Tuscany, Italy: IEEE, 2009: 468-472. |
[6] | Mai V, Natasha D, Vahid T. On the primary exclusive region of cognitive networks[J]. IEEE Transactions on Wireless Communications, 2009, 8(7): 3380–3385. doi: 10.1109/TWC.2009.080454 |
[7] |
廖勇, 杨士中, 李平, 等. 融合QoS与负载均衡的基础服务集信道分配算法[J]. 电子与信息学报, 2012, 34(9): 2230–2235.
Liao Yong, Yang Shizhong, Li Ping, et al. Integrated QoS and load balance among basic service set for channel allocation algorithm[J]. Journal of Electronics and Information Technology, 2012, 34(9): 2230–2235. |
[8] | Mi Penghui, Wang Xianbin. Improved channel assignment for WLANs by exploiting partially overlapped channels with novel CIR-based user number estimation[C]//2012 IEEE International Conference on Communications, Ottawa, Canada: IEEE, 2012: 6591-6595. |
[9] | Maziar N. A survey of cognitive radio access to tv spaces[C]//ICUMT'09. St. Petersburg, Russia: Hindawi Publishing Corporation, 2009: 1-8. |