TSCA：大规模无线可充电传感器网络的实时充电策略
52 浏览量
更新于2024-08-26
收藏 768KB PDF 举报
"TSCA：一种无线可充电传感器网络中按需架构的时空实时充电调度算法，旨在解决大规模无线可充电传感器网络（WRSNs）中的协同充电问题。该算法考虑了网络拓扑变化和节点故障等非确定性因素，以延长网络寿命。"
在无线可充电传感器网络（Wireless Rechargeable Sensor Networks，简称WRSNs）中，传感器节点通过无线充电车辆（Wireless Charging Vehicles，简称WCVs）进行能量补给，这为延长网络生命周期提供了新的思路。然而，现有的协同充电技术通常采用周期性和确定性的方法，未能充分考虑到如网络拓扑变化和节点故障等非确定性因素的影响，这使得这些方法在大规模WRSNs中应用时效果不佳。
本文提出了一种名为TSCA（Temporal-Spatial Real-Time Charging Scheduling Algorithm）的时空实时充电调度算法，特别针对按需充电架构设计。TSCA的主要目标是在尽可能减少死节点数量的同时，最大化能源效率以延长网络的使用寿命。
首先，WCV会收集各个传感器节点的充电请求。然后，它会计算一个可行的移动路径和充电策略，这个策略需要考虑到时间和空间的优化，以确保在处理充电请求的同时，有效地覆盖网络中的关键区域。为了实现这一目标，TSCA算法可能包含以下几个关键步骤：
1. **需求分析**：收集和分析各个传感器节点的剩余电量信息，预测其即将耗尽电量的时间，以及对网络功能的重要性。
2. **拓扑分析**：监测网络拓扑的变化，如新加入的节点、故障节点或移动节点，以便调整充电路线和时间。
3. **路径规划**：基于当前网络状态和充电需求，设计WCV的最优路径，以覆盖最多需要充电的节点，同时避免重复或无效的充电行为。
4. **实时调度**：根据节点的实时电量变化和网络状态，动态调整充电计划，确保在满足节点需求的同时，提高充电效率。
5. **能量效率优化**：通过智能算法，如遗传算法、模拟退火或深度学习等，寻找最佳的充电策略，以最小化WCV的能耗和时间，同时最大化网络的整体能量效率。
6. **故障恢复策略**：当遇到节点故障时，TSCA应能快速响应，重新分配资源，确保网络的关键功能不受影响。
TSCA算法的创新之处在于它考虑了非确定性因素，提高了充电调度的灵活性和适应性，这对于大规模、复杂环境下的WRSNs至关重要。通过这种方式，TSCA能够确保网络的稳定运行，降低维护成本，提高整体性能，从而在物联网、环境监测、智能交通等领域展现出广泛的应用潜力。
Charging Algorithm based on Task Splitting) are proposed
which aim at shortening charging time and distance via merg-
ing and splitting charging missions. In [9], we proposed a
Double Warning thresholds with Double Preemption
(DWDP) charging scheme, in which double warning thresh-
olds are used for initiating different kinds of charging
requests. In [4], a TemporAl and Distantial Priority charging
scheduling algorithm TADP is proposed for WRSNs. TADP
merges temporal priority and distantial priority into a mixed
priority to achieve better scheduling performance. In [28], a
game theoretical charging scheduling scheme is developed in
which WCVs collaboratively replenish nodes and the corre-
sponding situation is proved to reach a Nash Equilibrium.
Regardless of the fact that on-demand charging schemes
can indeed effectively solve the impact of uncertainty factors
in WRSNs, they still face some prominent problems such as
lacking of global optimum and underestimating the impacts
of low efficient nodes, which are urgent to be solved.
In this paper, our aim is to develop a charging schedul-
ing algorithm for the on-demand charging architecture of
WRSNs, which focuses on minimizing the number of
dead nodes and maximizing the energy efficiency from a
global view.
3PRELIMINARIES
In this section, related backgrounds are given.
3.1 Symbols and Definitions
We summarize the notations used in this paper in Table 1.
3.2 One Prominent Problem
A prominent and critical problem exists in this research
field is that the shortest Hamiltonian cycle is always
regarded as an optimal solution in nearly all scheduling
schemes [14], [29]. Their common objective is to minimize
the portion of traveling cost. As the traveling cost directly
relates to the length of the charging cycle, hence, a shortest
Hamiltonian cycle is widely regarded as the best and opti-
mal movement solution.
However, taking a deep investigation on temporal-
spatial characteristic of the charging tasks, we realize that,
usually the shortest Hamiltonian cycle is indeed the best
solution with the least energy consumption. But in some
extreme cases, it is not even a feasible solution. Surprisingly,
a common solution, which yields to neither the shortest
traveling cycle nor least charging energy cost, may become
a feasible and best solution for a WCV to choose. We exem-
plify this extreme case in Fig. 1.
As shown in Fig. 1, the charging request from node c is
the most urgent, which should be responded immediately.
The solution in Fig. 1a (a b c d e f g a)is
obtained by shortest Hamiltonian cycle, whose traveling
length and traveling cost are minimum, and the cycle in
Fig. 1b (a c b d e f g a) is a common solution.
When using the shortest Hamiltonian cycle as a solution,
node c cannot get charged before its deadline and become
dead. Although it is an optimal solution for traveling, it is
not even a feasible solution for charging scheduling. On the
contrary, a solution (a c b d e f g a) with big-
ger traveling length and charging cost can meet the tempo-
ral, spatial, and energy requirements of node c. It can be
regarded as a feasible solution or even the best solution. In
the following work, we pay close attention to a) how to find
a feasible solution for charging scheduling and b) how to
optimize it.
3.3 Problem Statement
We envision a scenario that in a WRSN, a number of
homogenous sensor nodes are randomly deployed in an
open environment shown in Fig. 2. One WCV is responsible
for replenishing energy for all the sensors. When the
TABLE 1
Symbols and Definitions
Symbol Definition
T The current time.
t
i
The arrival time when WCV reaches node i.
N
task
The number of completed charging tasks.
p
i
The probability of status i.
N The total number of nodes in the sensor network.
P A charging path.
p
i;j
The charging path from node i to node j.
D
i;j
The distance between i and j.
E
C
ðiÞ The remaining energy of node i.
P
T
ðiÞ The temporal priority of node i.
P
M
The battery energy of sensor.
u The threshold for sending a charging request.
T
R
ðiÞ The remaining time of node i.
T
D
ðiÞ The deadline (or exhaustion time ) of node i.
T
m
The average time of charging one single node.
S
N
ðiÞ The work state of a node i (alive or dead).
M
A
ðPÞ The minimum remaining time of a node in a path P.
T
C
ðiÞ The increased time after node i is inserted into the path.
E
e
The energy amount finally received by nodes.
h The energy efficiency of WCV.
Q
M
ðjÞ A mission queue of task j that contains requests
to be responded, jQ
M
ðjÞj indicates its size
1
.
Q
W
A task queue that records all requests which
are not responded.
Q
P
A queue recording the nodes which are to
be dropped.
Q
B
ðjÞ A feasible charging queue for the task j.
Q
H
ðjÞ A charging queue without low efficient nodes
for the task j.
Q
optimized
ðjÞ A charging queue that has been optimized
for the task j.
The actual average queue length of requests.
V
M
ðiÞ The energy consumption rate of node i.
V
D
The energy consumption of WCV in a unit
distance.
v The speed of WCV.
N
D
The set of dead nodes.
r The service intensity of WCV.
The average strength of charging request
which follows a Poisson distribution.
m The average service time of the WCV.
s
2
The variance of WCV’s service time.
X
n
The number of nodes in Q
W
and Q
M
when
node n has been charged.
T
n
The serving time that node n þ 1 needs when
node n has been charged.
Y
n
The number of nodes requesting for charging
when node n þ 1 is being served.
L
s
The average number of nodes that are
waiting for getting charged in Q
W
and Q
M
.
L
q
The average number of nodes in Q
W
.
W
s
The average sojourning time of a charging
request in Q
M
.
W
q
The average waiting time of a node to get
charged in Q
M
.
LIN ET AL.: TSCA: A TEMPORAL-SPATIAL REAL-TIME CHARGING SCHEDULING ALGORITHM FOR ON-DEMAND ARCHITECTURE IN... 213
剩余13页未读，继续阅读
2021-02-08 上传
2010-03-12 上传
2021-04-28 上传
2021-09-30 上传
2022-01-07 上传
2013-10-30 上传
2021-02-21 上传
2021-10-18 上传
2009-09-22 上传
weixin_38617001
- 粉丝: 5
- 资源: 902
我的内容管理
展开
最新资源
- IEEE 14总线系统Simulink模型开发指南与案例研究
- STLinkV2.J16.S4固件更新与应用指南
- Java并发处理的实用示例分析
- Linux下简化部署与日志查看的Shell脚本工具
- Maven增量编译技术详解及应用示例
- MyEclipse 2021.5.24a最新版本发布
- Indore探索前端代码库使用指南与开发环境搭建
- 电子技术基础数字部分PPT课件第六版康华光
- MySQL 8.0.25版本可视化安装包详细介绍
- 易语言实现主流搜索引擎快速集成
- 使用asyncio-sse包装器实现服务器事件推送简易指南
- Java高级开发工程师面试要点总结
- R语言项目ClearningData-Proj1的数据处理
- VFP成本费用计算系统源码及论文全面解析
- Qt5与C++打造书籍管理系统教程
- React 应用入门：开发、测试及生产部署教程
