1816 IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT, VOL. 58, NO. 6,JUNE 2009 Remaining Capacity Measurement and Analysis of Alkaline Batteries for Wireless Sensor Nodes Adriano B. da Cunha, Breno R. de Almeida, and Diógenes C. da Silva, Jr., Member, IEEE Abstract—In most applications, wireless sensor networks (WSNs) will deploy a large number of distributed sensor nodes in remote or inhospitable places, making batteries their main source of energy; thus, the stored energy is a key resource of a WSN. Sensor nodes should balance their limited resources to increase the lifetime of the network. The knowledge of the available amount of energy becomes an important requirement for the maintenance, implementation of self-management techniques, and viability of the WSN. Therefore, the research of the State-of-Charge (SoC), or the remaining capacity estimation, is of key importance. This paper presents an energy-efficient battery-remaining capacity- estimation technique. The experiments were conducted using the MICA2 wireless sensor node platform, which shows that the voltage-only-based estimation presented an available 18% of the battery maximum capacity, although the battery had been fully discharged, and a current-based estimation technique is presented with minimal hardware intervention. Index Terms—Battery models, battery remaining capacity, State-of-Charge (SoC) of batteries, wireless sensor networks (WSNs), wireless sensor nodes platforms. I. I NTRODUCTION A WIRELESS sensor network (WSN) is composed of many autonomous and compact devices called sensor nodes. They have limited resources, such as computational capacity, memory, communication, and energy. In most applications, WSNs will deploy a large number of distributed sensor nodes in remote or inhospitable places, making batteries their main source of energy. While computer processor speed and memory capacity have exhibited magnitude orders of increase since 1990s, the battery energy density has only tripled, that is, the energy supply and storage technology is not advancing at the same rate [1]. As the network lifetime depends on the amount of available energy, each sensor node should balance its limited resources to increase the overall network lifetime. Keeping the sensor nodes properly operating and for as long as possible using battery-limited energy has become a critical issue. Moreover, the necessity to manage the network resources, mainly the available energy, in an intelligent and autonomous way becomes an important requirement for the maintenance, Manuscript received January 25, 2008; revised September 24, 2008. First published February 27, 2009; current version published May 13, 2009. This work was supported in part by Fundação de Amparo à Pesquisa do Estado de Minas Gerais, Conselho Nacional de Desenvolvimento Científico e Tec- nológico, and Fundação de Desenvolvimento da Pesquisa/Universidade Federal de Minas Gerais, Belo Horizonte. The Associate Editor coordinating the review process for this paper was Dr. Flavio Vasconcelos. The authors are with the Federal University of Minas Gerais, Belo Horizonte, MG-31.270-010 Brazil. Color versions of one or more of the figures in this paper are available online at http://ieeexplore.ieee.org. Digital Object Identifier 10.1109/TIM.2009.2013660 implementation of self-management techniques, and viability of the WSN. That is why the research of the State-of-Charge (SoC), or the remaining capacity estimation, is of key impor- tance. The objective is to provide mechanisms inside the WSN to determine the lifetime for each sensor node. One possible solution is to estimate the battery remaining capacity. However, traditional techniques for the battery remaining capacity esti- mation do not adjust to WSN, and new alternatives must be studied and analyzed. The objective of this paper is to present aspects to be consid- ered for the battery remaining capacity estimation for wireless sensor nodes. A reduced and initial version of this paper was presented at the Seventh International Seminar on Electrical Metrology (VII SEMETRO) in 2007 [2]. This paper extends and improves the earlier proposed battery model and capacity- estimation method. To obtain the necessary information, a minimal hardware device is presented, which is called current measurement device (CMD). The remainder of this paper is organized as follows. Section II presents a quick tutorial on WSNs. Section III presents batteries with the concepts and the factors that influence the batteries’ capacity, the main methods for the battery remaining capacity estimation, and the battery models found in the literature. Section IV presents the exper- iments and results obtained using a MICA2 sensor node from Mica Motes platform [3] and a 14-bit data acquisition board (DAQ) (NI USB-6009) from National Instruments [4]. Finally, the conclusions are presented in Section V. II. WSN ARCHITECTURE This section gives an overview of the WSN architecture. WSNs are networks composed of a large number of sensor nodes. The objective of these networks is to collect data. Sensor nodes are usually deployed over a desired area; then, they wake up, self test, and establish dynamic communications among them, composing a network [5]. Each sensor node senses the environment, processes (e.g., fusion, aggregation, and others), and usually transmits the data to gateway nodes using a multihop communication. Gateway nodes transmit their data to an external observer called base sta- tion. Gateway nodes are ordinary sensor nodes or more complex devices having more computational capabilities like greater radio range and more computational power. The discussion of gateway nodes is outside the scope of this paper. In a conventional network, such as cellular phone or local wireless networks, the communication between computational elements is done through radio base stations, which represent a communication infrastructure. WSNs usually do not have 0018-9456/$25.00 © 2009 IEEE