2型模糊逻辑与蚁群优化驱动的WSN均衡路由协议

173 浏览量更新于2024-07-15 收藏 1.31MB PDF 举报

本文主要探讨的是无线传感器网络（WSN）中的路由优化问题，特别是在考虑网络负载均衡和延长网络使用寿命方面的挑战。针对这一问题，研究人员提出了一个名为CRT2FLACO的基于2型模糊逻辑（Type-2 Fuzzy Logic）和蚁群优化（Ant Colony Optimization, ACO）的WSN集群路由协议。在研究中，2型Mamdani模糊逻辑系统（T2MFLS）起到了关键作用。T2MFLS被设计用来处理WSN中复杂多变的不确定性，如节点的剩余能量、邻居数量以及节点到基站（BS）的距离等。这三个因素作为输入，通过T2MFLS的处理，确定每个节点成为候选簇头（Cluster Head, CH）的概率以及CH的竞争半径。这个过程旨在找到最优的CH，确保网络负载均衡，减少通信压力。在集群建立阶段，通过T2MFLS的选择机制，能够动态地选择具有适宜能力和资源的节点作为CH，有效防止过早耗尽能源的节点成为负担。这有助于在网络生命周期早期就开始实施负载均衡策略，提高整体效率。进入稳态阶段后，所有CH通过蚁群算法形成一个链状结构，每个CH将数据包沿着这条链传递给最终的领导者，也就是负责将数据包进一步转发给BS的CH。这种分层次的数据传输方式不仅减少了数据包在单个节点间的传输次数，降低了传输能耗，还使得数据路径更加高效，从而显著延长了WSN的使用寿命。通过仿真结果，研究者证明了CRT2FLACO协议的有效性。它不仅在理论上解决了WSN的负载均衡问题，而且在实际应用中也展示了优秀的性能，是当前无线传感器网络中一个值得重视的路由优化解决方案。这项工作对于提升无线传感器网络的可靠性和效率，特别是在资源受限和环境不确定的环境中，具有重要的理论和实践价值。

exhaustion of energy for some nodes maybe cause the disconnection of a WSN, and

consequently could not complete a task. One should select the nodes with more energy to

be CHs. Second, the node density should also be considered, since if we choose a node

with large number of neighbor nodes to be a CH, this node will consume more energy for

data fusion with its cluster and data transmission to BS, and then die fast. At last, the nodes

far from BS will also consume more energy for data transmission to BS, and then die fast.

In view of the above, we take into consideration residual energy, the number of neighbor

nodes and the distance to BS of each node for CHs election. And in order to avoid CHs

gathering, we also introduce a strategy of unequal competition radius.

Before the sensor network begins to collect data, the BS broadcasts a control signal of

ﬁxed power level to all sensor nodes. According to the strength of the received signal,

sensor nodes can calculate the distance to BS, then send the distance and residual energy

information to BS. Similarly, at the beginning of each round, each node reports its residual

energy to BS. Based on the received information the BS simulates sensor network and

applies the proposed T2FLCA algorithm to calculate the CHs list, then generates and sends

control package to each sensor node, and each ordinary node determines which cluster it

can join in based on the distance factor, and sends control package and data packet to their

CHs. Finally all the CHs are linked into a chain using ACO algorithm, each CH send fusion

data packet to the leader along link, and the leader transmits eventually packets to BS.

4.1 Architecture of a Type-2 Fuzzy Logic System

In this paper, a type-2 Mamdani FLS (T2MFLS) is established by T2FLCA algorithm,

including four parts [14]: Fuzziﬁer, Rules, Inference Engine, Output processing (Type-

reducer and Defuzziﬁer), as shown in Fig. 1.

In the algorithm, the input variables include the residual energy, the number of neighbor

nodes and the distance to BS of each node; two output variables are the probability of a

node becoming the candidate CH and the competition radius of each candidate CH.

4.1.1 Fuzziﬁer and Division of Fuzzy Variables

The role of fuzziﬁer is mapping each crisp input value to a fuzzy set. For a T2MFLS, the

fuzziﬁer maps each crisp input vector x

¼ðx

; ...; x

Þ to a type-2 fuzzy set

. In order to

simplify the calculation, we use singleton fuzziﬁer, that is, for 8i ¼ 1; ...; p, the mem-

bership function of the fuzziﬁed input set

is deﬁned as [14]:

Fig. 1 The structure of a type-2 Mamdani fuzzy logic system

1170 W.-X. Xie et al.

123

ðx

Þ¼

1=1 x

¼ x

1=0 x

6¼ x



ð4Þ

and l

ðxÞ¼u

i¼1

ðx

Þ, where x ¼ðx

; ...; x

Þ, and the notation u denotes the operation

of intersection of type-2 fuzzy sets.

For convenience, the domains of three inputs are all taken to be the unit interval [0,1].

We get the input values of the residual energy, the number of neighbor nodes and the

distance to BS by using the corresponding real values divided by the initial energy,

maximum number of neighbor nodes and maximum distance, respectively. To reduce the

computation complexity, we divide all the three inputs into three levels, that is, the residual

energy (denoted by Energy): low, medium, high; the number of neighbor nodes (denoted by

Density): sparse, medium, dense; the distance to BS (denoted by Distance): near, medium,

far. All secondary membership functions (MFs) are taken to be interval sets (this means

our FLS are interval type-2), and the primary memberships are triangular or trapezoid (see

[14] for the concepts of secondary MFs and primary memberships). Figures 2, 3 and 4

show the three inputs and their membership functions.

The two output variables are the probability of the node becoming the candidate CH

(denoted by Probability) and the competition radius of each candidate CH (denoted by

Radius). Their domains are the real values. They are divided into nine levels: very weak,

weak, little weak, lower medium, medium, higher medium, little strong, strong, very strong

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

low medium high

Energy

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

0.2

0.4

0.6

0.8

Energy

Membership grade

highmediumlow

(a) (b)

Fig. 2 T2 and T1 membership functions of input variable ‘Energy’. a T2 MFs, b T1 MFs

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

sparse medium dense

Density

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1

0.2

0.4

0.6

0.8

Density

Membership grade

densemediumsparse

(a) (b)

Fig. 3 T2 and T1 membership functions of input variable ‘Density’. a T2 MFs, b T1 Mfs

A Clustering Routing Protocol for WSN Based on Type-2 Fuzzy… 1171

123

剩余31页未读，继续阅读

weixin_38638033

粉丝: 5
资源: 940

2型模糊逻辑与蚁群优化驱动的WSN均衡路由协议

基于蚁群算法的WSN路由优化：解决LEACH问题

改进蚁群算法在WSN路由优化研究

蚁群算法优化的WSN能量预测路由协议

基于混沌粒子群与蚁群算法的WSN路由协议

Matlab基于蚁群算法的wsn路由设计-一种基于蚁群算法的WSN路由算法.pdf

论文研究-基于改进蚁群算法的WSN路由协议 .pdf

Matlab基于蚁群算法的wsn路由设计-一种基于蚁群算法的能量有效wsn分簇算法.pdf

vrrya.rar_WSN_WSN覆盖优化_WSN覆盖算法_蚁群算法 WSN_蚁群算法WSN

基于改进蚁群算法的WSN路由研究.pdf

论文研究-一种基于蚁群算法的WSN路由算法.pdf

最新资源