Sizing up the Batteries Modelling of Energy-Harvesting Sensor Nodes in a Delay Tolerant Network

2025-05-03 0 0 484.92KB 13 页 10玖币

侵权投诉

Sizing up the Batteries: Modelling of

Energy-Harvesting Sensor Nodes in a Delay

Tolerant Network

Jeremiah D. Deng

Abstract For energy-harvesting sensor nodes, rechargeable batteries play a critical

role in sensing and transmissions. By coupling two simple Markovian queue models

in a delay-tolerant networking setting, we consider the problem of battery sizing for

these sensor nodes to operate eﬀectively: given the intended energy depletion and

overﬂow probabilities, how to decide the minimal battery capacity that is required to

ensure opportunistic data exchange despite the inherent intermittency of renewable

energy generation.

1 Introduction

Recently, energy-harvesting wireless sensor networks (EH-WSN) [1] have become

a promising technology for sensing applications. The advantage of EH-WSN is

obvious - batteries on the sensor nodes can be downsized due to their energy-

harvesting capability, the network enjoys longer life time, eliminating the need

of frequent of battery replacement, which is especially challenging for large-scale

sensor deployment. However, apart from reservoirs, most renewable energy sources

are intermittent in nature, which raises new challenges in designing EH-WSNs. For

example, sensors may not get proper sunshine for recharging for hours, and wearable

devices operated by kinetic energy will not beneﬁt much from humans sitting for

hours. This implies the necessity of using batteries to buﬀer the unsteady power

supply from renewable energy sources.

We consider a generic EH-WSN scenario where mobile nodes are equipped

with capacity-limited batteries that are powered by harvested kinetic energy; data

exchange between nodes requires 1) they are within transmission range to each other;

and 2) there is suﬃcient energy to conduct data transmission. This is in eﬀect an

Jeremiah D. Deng

Department of Information Science, University of Otago, PO Box 56, Dunedin 9054, e-mail:

jeremiah.deng@otago.ac.nz

arXiv:2210.05316v1 [cs.NI] 11 Oct 2022

2 Jeremiah D. Deng

EH-WSN operating as a delay-tolerant network (DTN) [15], where data transmission

is opportunistic. In such a scenario, it is both important to ensure the battery size is

large enough to avoid energy depletion (and hence potential failure for transmission)

and energy overﬂow, both detrimental to the battery life.

In a previous work [17], we have examined battery sizing in terms of depletion

probability and overﬂow probability respectively, using a coupled data and energy

queue system. In this work, we intend to investigate the mathematical properties

of battery size as a function regarding the operational probability requirement, and

develop an algorithm to calculate the minimum battery size needed to meet the given

requirements.

2 Related Work

As a performance modelling tool, queueing theory has been employed to study

EH-WSNs. Gelenbe [5] ﬁrst looked the modelling of an EH-sensor node using the

concept of discretized energy unit called “energy packets”. The arrival of these

energy packets is assumed to follow a Poisson process. A routing approach was

further developed in [6]. A more general queueing model was introduced in [10],

relaxing the assumption that exactly one energy packet is required to transmit a data

packet. A Markovian model with data buﬀering was further considered in [4]. In a

recent work [17] we showed that kinetic energy harvested by ﬁtness gears discretized

as energy packets can be well modelled by Poisson processes. These previous works,

however, considered only static EH sensors, without involving potential intermittent

connections between EH-sensor nodes due to mobility.

On the other hand, mobility has been widely investigated in ordinary wireless

sensor networks and DTNs [12, 14]. Despite some counter-arguments [3], several

mobility model studies [2, 8, 16, 12] suggested that two mobile nodes’ encounter

follows a Poisson process in mobile ad hoc networks and DTNs. There are few

studies on energy harvesting networks that investigated the eﬀects of intermittent

connections [11, 13].

3 System Modelling

Notations used in this article are listed as follows:

𝜆𝐸energy packet arrival rate

𝜆𝐷data packet arrival rate

𝜆𝐶connection arrival rate

𝛾𝐷ratio 𝜆𝐷/𝜆𝐶

𝛾𝐸ratio 𝜆𝐸/𝜆𝐶

𝛾ratio 𝜆𝐷/𝜆𝐸

𝑃𝐷0proportion of time that there is no data in the system

Sizing Up The Batteries 3

𝑃𝐸𝑘proportion of time that system have 𝑘energy packets 𝑘=0, ..., 𝐾

𝜌𝐷utilization factor of data buﬀer

𝜌𝐸utilization factor of energy buﬀer

𝛼acceptable probability of energy depletion

𝛽acceptable probability of energy overﬂow

𝐾𝛼battery capacity decided based on 𝛼

𝐾𝛽battery capacity decided based on 𝛽

d𝑥eceiling, the greatest integer more than or equal to 𝑥

3.1 The queueing model

We consider a network of mobile EH-sensors. Energy harvesting leads to Poisson

arrivals of energy packets (EP) with a rate of 𝜆𝐸. Energy consumption occurs when

there are data packets in buﬀer, provided that there are nodes in proximity, which is

modulated by another Poisson arrival rate 𝜆𝐶. Thus an Energy queue is formed at

each sensor node, which can be modelled as an 𝑀/𝑀/1/𝐾, where 𝐾is the battery

capacity (in terms of number of energy packets). Data packets (DP) arrive at a

Poisson rate 𝜆𝐷, and leave a node if there is a connection available and there is at

least an energy packet in system. As memory in a sensor node is relatively cheap and

less constrained, for simplicity we set no limit to the data buﬀer, hence allowing the

data queue to be modelled by an 𝑀/𝑀/1. Clearly both queues are coupled by the

connection availability. Hence the Markovian packet departures in both the Energy

queue and the Data queue are modulated by the connection arrival rate 𝜆𝐶.

The system diagram for a sensor node is shown in Figure 1.

Data Buffer

Battery

Connection

Fig. 1: System diagram of a mobile sensor node. DPs arrive in the data buﬀer, while

EPs arrive in the battery. Consumption of the energy as well as the transmission of

data occur simultaneously when triggered by a connection established with another

node.

Similar to [10] and [6], we focus on modelling energy needed for data transmission

and assume that compared with data transmission the sensing process consumes

insigniﬁcant amount of energy from the battery.

It is also worth mentioning that data transmission time is much faster than energy

harvesting in a node. Given the size of sensory DP and the relatively large bandwidth

文档加载中……请稍候！
如果长时间未打开，您也可以点击刷新试试。

下载文档到电脑，查找使用更方便

10 玖币 0人已下载

立即下载

摘要：

SizinguptheBatteries:ModellingofEnergy-HarvestingSensorNodesinaDelayTolerantNetworkJeremiahD.DengAbstractForenergy-harvestingsensornodes,rechargeablebatteriesplayacriticalroleinsensingandtransmissions.BycouplingtwosimpleMarkovianqueuemodelsinadelay-tolerantnetworkingsetting,weconsidertheproblemofbat...

展开>> 收起<<

Sizing up the Batteries Modelling of Energy-Harvesting Sensor Nodes in a Delay Tolerant Network.pdf

共13页,预览3页

还剩页未读，继续阅读

声明：本站为文档C2C交易模式，即用户上传的文档直接被用户下载，本站只是中间服务平台，本站所有文档下载所得的收益归上传人(含作者)所有。玖贝云文库仅提供信息存储空间，仅对用户上传内容的表现方式做保护处理，对上载内容本身不做任何修改或编辑。若文档所含内容侵犯了您的版权或隐私，请立即通知玖贝云文库，我们立即给予删除！

Sizing up the Batteries Modelling of Energy-Harvesting Sensor Nodes in a Delay Tolerant Network

相关推荐

开通VIP享超值会员特权

作者详情

相关内容

热门标签

举报选择: