Elsevier

Computer Communications

Volume 32, Issue 5, 27 March 2009, Pages 896-906
Computer Communications

Application-driven, energy-efficient communication in wireless sensor networks

https://doi.org/10.1016/j.comcom.2008.12.022Get rights and content

Abstract

Several sensor network applications based on data diffusion and data management can determine the communication transfer rate between two sensors beforehand. In this framework, we consider the problem of energy efficient communication among nodes of a wireless sensor network and propose an application-driven approach that minimizes radio activity intervals and prolongs network lifetime. On the basis of possible communication delays we estimate packet arrival intervals at any intermediate hop of a fixed-rate data path. We study a generic strategy of radio activity minimization wherein each node maintains the radio switched on just in the expected packet arrival intervals and guarantees low communication latency. We define a probabilistic model that allows the evaluation of the packet loss probability that results from the reduced radio activity. The model can be used to optimally choose the radio activity intervals that achieve a certain probability of successful packet delivery for a specific radio activity strategy. Relying on the probabilistic model we also define a cost model that estimates the energy consumption of the proposed strategies, under specific settings. We propose three specific strategies and numerically evaluate the associated costs. We finally validate our work with a simulation made with TOSSIM (the Berkeley motes’ simulator). The simulation results confirm the validity of the approach and the accuracy of the analytic models.

Introduction

A wireless sensor network is a specialized ad hoc network composed of a large number of low power, low cost nodes (also called sensors) [1]. Each node in a sensor network is equipped with a processor, one or more sensing units, a radio transceiver, and is powered with an embedded battery.

Sensors are spread in the surrounding environment (sensor field) without any predetermined infrastructure and cooperate to execute common monitoring tasks. They self-organize into a wireless ad hoc network in order to exchange sensed data and to connect with external sink nodes that issue queries to the network.

Habitat and environmental monitoring represent a class of sensor network applications that have enormous potential benefit for the entire scientific community [2], [3], [4]. These applications share a common structure, where fields of sensors are tasked to take periodic readings, analyze and pre-process data, and report the analysis to external sink nodes.

Due to the lack of communication infrastructure and the limited battery power, the radio range of a node is limited, thus it can communicate directly only with neighbouring nodes. Communication with other sensors is generally multihop: all sensors cooperate to forward messages to their destinations using a routing protocol [5]. The sensor network can be programmed according to different paradigms which define the communication model among the sensors and between the sensor network and the sinks. In the data diffusion paradigm [6] the network is organized into a directed acyclic graph rooted at the sink, and sensed data flow with different rates from sensors to the sink. In recent paradigms [7], [8] the network is seen as a database and the sink programs the network by sending queries to sensors.

Efficient use of energy is of fundamental importance in sensor networks [9] and several techniques have been proposed in order to prolong battery lifetime by exploiting energy-efficient communication strategies.

The principal sources of power consumption are collisions and retransmissions at the MAC layer, overhearing of packets destined to other nodes, control packet overhead and idle listening, i.e. maintaining the radio in idle mode in order to listen continuously for possible incoming transmissions [10]. Measurements have shown that idle listening consumes 50–100% of the energy required for receiving [9], [11]. A possible approach to reduce the idle intervals is to turn off the radio interface, therefore saving a considerable amount of energy. Clearly, if a source node sends some packets to a destination node, some form of synchronization between the sender and the receiver must be introduced in order to properly schedule the activation of their radio.

Most of the proposed schemes typically develop scheduling algorithms at the MAC layer or exploit coordinated activities of the network and MAC layers. These approaches reduce energy consumption, however, the information available at the MAC layer (and possibly at the network layer) used to schedule communications are not sufficient and in several cases many nodes switch on their interface unnecessarily or for longer periods than needed. Thus, to properly reduce the power consumption a strict layered networking design is insufficient and research studies on cross-layer designs are particularly active [12], [13]. In this paper we consider an energy-efficient communication model for sensor networks which is driven by the application. Using information from the application layer may dramatically improve power consumption.

The application driven model, is suitable especially for applications where sensors communicate by sending data streams to each other (as it is the case in data diffusion), with regular rate. It exploits information from the application layer to synchronize the sensors along a communication channel in order to forward the data stream to the destination.

In our model the time axis is divided into periods. Each period comprises a system window in which the application constructs channels and a communication window. In the communication window only sensors belonging to a communication channel are active. These nodes are synchronized, exploiting application driven information, in order to forward data streams to their destination with the minimum possible latency. In system windows all the sensors are active in order to provide connectivity between an arbitrary pair of sensors. Note that MAC layer solutions can be used during system windows to further reduce power consumption.

We propose three strategies to coordinate the sensors in a certain communication channel in order to reduce the time they should have the radio on. Focusing on a communication channel, we give a probabilistic model under which the probability of packet loss can be analytically evaluated and we provide a cost model to evaluate the energy efficiency of the three strategies. Then we discuss how the strategies can be tuned according to different requirements of the application.

The rest of the paper is organized as follows. Section 2 reviews related work and in Section 3 we present an applicative scenario. Section 4 describes the communication protocol and introduces a general framework for the analysis. Section 5 develops a probabilistic and cost model for the general framework and Section 6 studies the radio activity optimization in the probabilistic model. Section 7 proposes alternative strategies for the optimization. Simulation results appear in Section 8 and conclusions are drawn in Section 9.

Section snippets

Related work

Among the MAC layer solutions, [14], [15], and [16] require that each sensor be equipped with an additional radio, reserved for signalling, which is used to synchronize communication with the main radio. MAC solutions exploiting synchronization protocols can be found in [17], [18], [19]. In these protocols sensors divide their time axis into frames. Each frame is divided into an active window (during which the radio is on) and a power-save window (in which the radio is off). Typically the size

Scenario

There are several applications of wireless sensor networks where it is possible determine in advance the time intervals where pairs, or even groups, of nodes need to communicate. In these cases, as proposed and analyzed in this paper, it is possible to apply synchronization strategies that require nodes to activate their interfaces just when and where really needed.

As a simple example, consider the case of a node u1 of wireless sensor network that has been instructed to acquire information

Communication paradigm

As we discussed in Section 3, in several cases it is possible to know in advance when applications need to send data through the sensor network and, typically, sensor networks applications request data at specific rates from specific nodes.

This knowledge can be exploited in order to reduce energy consumption in single-hop communications. In this case it is sufficient that the sender and receiver be synchronized and turns on their radio interface simultaneously, at application dependent

Probabilistic model

As discussed in Section 4, every node may incur some delays before being able to actually send a packet. Internal processing and, chiefly, contention for access to the shared medium are the causes of these delays. We model the total delay introduced by node ui, i[0,n-1], with a continuous random variable Di and density function fi. In order to simplify the analysis, we assume that the Dis are identically distributed independent random variables. For m[1,n], we also define:Rm=i=0mDithe total

Optimization of the cost function

Our approach, which maintains the radio in idle mode for a wait time w(i) in the nodes of a channel, is able to reduce energy consumption when either a packet is lost during transmission, or when a packet is not sent (for various application dependent reasons) at the transmission time. It does not contribute to save energy in case of successfully delivering a packet to destination. In fact, our transmission protocol, described in Section 4, switches the radio off as soon as a message is

Alternative strategies for w(i)

It is clear from the discussion of the previous section that solving the optimization problem in the sensors is, in general, rather expensive in terms of processing time and energy. For this reason, in a real deployment, we envision two options to achieve this assessment:

  • 1.

    All decisions are taken by the sink node, which asks node u0 to open a channel toward node un. In this case with the knowledge of the network statistics, updated time after time, the sink computes the optimal parameters to be

Simulations

We implemented our strategies on the nesC/TinyOs platform [28,29] and ran simulations with TOSSIM [30], the simulator for mica motes [31]. Our goal was to validate the probabilistic model we developed in Section 5. We had to modify (in straightforward ways) some of the TinyOs modules in order to obtain millisecond accuracy in the measurements.

We set up a channel of 8 nodes and simulated a fixed data rate using packets with a 20-byte payload. We measured the number of packets lost and the medium

Conclusions

We considered an application-driven communication paradigm for sensor networks. The paradigm exploits information from the application layer in order to set up channels between pairs of sensors, along which a data stream travels with rates defined by the application. Focusing on a single channel, we developed a probabilistic model under which the probability of packet loss can be analytically evaluated and we gave a cost model to evaluate the energy efficiency. We proposed three specific radio

References (28)

  • Paolo Baronti et al.

    Wireless sensor networks: a survey on the state of the art and the 802.15.4 and ZigBee standards

    Computer Communications

    (2007)
  • R. Szewczyk et al.

    Habitat monitoring with sensor networks

    Communications of the ACM

    (2004)
  • A. Mainwaring, D. Culler, J. Polastre, R. Szewczyk, J. Anderson, Wireless sensor networks for habitat monitoring, in:...
  • R. Szewczyk, A. Mainwaring, J. Polastre, J. Anderson, D. Culler, An analysis of a large scale habitat monitoring...
  • B. Karp, H.T. Kung, GPSR: greedy perimeter stateless routing for wireless networks, in: Proceedings of the...
  • C. Intanagonwiwat, R. Govindan, D. Estrin, Directed diffusion: a scalable and robust communication paradigm for sensor...
  • S. Madden, M.J. Franklin, J.M. Hellerstein, W. Hong, The design of an acquisitional query processor for sensor...
  • S. Madden, M.J. Franklin, J.M. Hellerstein, W. Hong, TAG: a tiny aggregation service for ad-hoc sensor networks, in:...
  • V. Raghunathan et al.

    Energy-aware wireless microsensor networks

    IEEE Signal Processing Magazine

    (2002)
  • W. Ye et al.

    Medium access control with coordinated adaptive sleeping for wireless sensor networks

    IEEE/ACM Transactions on Networking

    (2004)
  • M. Stemm et al.

    Measuring and reducing energy consumption of network interfaces in hand-held devices

    IEICE Transactions on Communications

    (1997)
  • R. Min, M. Bhardwaj, N. Ickes, A. Wang, A. Chandrakasan, The hardware and the network: total-system strategies for...
  • E. Shih et al.

    Design considerations for energy-efficient radios in wireless microsensor networks

    The Journal of VLSI Signal Processing-Systems for Signal Image and Video Technology

    (2004)
  • C. Schurgers, V. Tsiatsis, S. Ganeriwal, M. Srivastava, Topology management for sensor networks: exploiting latency and...
  • Cited by (14)

    • Optimized query routing trees for wireless sensor networks

      2011, Information Systems
      Citation Excerpt :

      Thus, the answer will be ready prior the completion of time instance 31 which is the end of the epoch. A recent paper that proposes a scheduling algorithm for wireless sensor networks has been presented in [3]. The authors define a probabilistic model that allows the evaluation of the packet loss probability that results from the reduced radio activity.

    • Resource Allocation in Wireless Powered Communication Networks with Power Minimization

      2022, 2022 International Conference for Advancement in Technology, ICONAT 2022
    • A testbed and an experimental public dataset for energy-harvested IoT solutions

      2019, IEEE International Conference on Industrial Informatics (INDIN)
    • Statistical Energy Neutrality in IoT Hybrid Energy-Harvesting Networks

      2018, Proceedings - IEEE Symposium on Computers and Communications
    View all citing articles on Scopus
    View full text