Throughput, Energy Consumption, and Energy ... - Semantic Scholar

7 downloads 2721 Views 363KB Size Report
802.15.6 WBAN MAC protocol in terms of throughput, power ... conditions, supervised recovery from a surgical procedure, and monitoring ... different data rates.
Throughput, Energy Consumption, and Energy Efficiency of IEEE 802.15.6 Body Area Network (BAN) MAC Protocol Byoung Hoon Jung, Raja Usman Akbar, and Dan Keun Sung Department of EE, KAIST, KOREA E-mail: [email protected], [email protected] Abstract—With increasing interest in e-health care, integration of low-powered and light-weight sensor nodes leads to the requirement and development of Wireless Body Area Networks (WBAN). Recently, the Institute of Electrical and Electronic Engineering (IEEE) introduced a new standard, IEEE 802.15.6 for wireless body area communications. The goal of this standard is to specify several physical layers (PHY) and medium access control (MAC) layer protocols for variety of applications with various QoS requirements. In this paper, we evaluate the performance of the IEEE 802.15.6 WBAN MAC protocol in terms of throughput, power consumption, and energy efficiency under unsaturated conditions. We develop a discrete-time Markov chain based analytical model to evaluate the performance of CSMA/CA based WBAN MAC protocol during contention access phases. To verify the numerical results obtained from analytical model, we performed simulations and compared the results.

I. I NTRODUCTION With exponentially increasing demands for communication services, various wireless networking technologies have been developed and deployed. The convergence issues among different research areas such as e-health services [1] have been of interest to scientists and researchers. Wireless sensors networks (WSNs) have been studied in a wide range of applications in health care. WSNs consist of intelligent sensor nodes that are capable of gathering and processing information in a given environment and send it to the remote base stations. Wireless Body Area Network (WBAN) is considered as a special type of WSN. It consists of low-power wireless sensors nodes, which are placed or implanted on or inside of a human body for continuous health monitoring. WBAN is a key technology to prevent the occurrence of myocardial infarction, monitor episodic events or any other abnormal conditions and can be used for ambulatory health monitoring. Additionally, it can also be used in diagnostic procedure, maintenance of chronic conditions, supervised recovery from a surgical procedure, and monitoring the effects of drugs therapy. Recently, the Institute of Electrical and Electronics Engineers (IEEE) has launched the Task Group (TG-6) which specifically defines and sets the specifications for WBAN. In recent years, WBANs have been studied in the literature. These studies are mostly focused on various technical issues of the WBANs. The IEEE has standardized various protocols to support different access networks, for example: WiMax (IEEE 802.16), WLAN (IEEE 802.11), and WPAN (IEEE

802.15). IEEE 802.15 Task Group 6 (BAN) is developing a communication standard optimized for low power devices and operation on, in or around the human body (but not limited to humans) to provide a variety of applications including medical, consumer electronics / personal entertainment and others [2]. They standardized IEEE P802.15.6/D01 Draft Standard for Body Area Network in May 2010 [3]. This draft is still under revision; however, it provides the details about the requirements for body area communication. Before introducing the IEEE 802.15.6 standard by the IEEE 802.15 Working Group, the structure of WBANs and protocols and mechanisms of the physical layer and MAC sub-layer of WBANs have been one of the most important concerns. In [4], the authors provided a comprehensive survey on Wireless Body Area Network. Performance analysis of any wireless access network is an important step to evaluate the system. Since WBAN nodes are powered by small batteries, energy consumption is also one of important issues in evaluating the system. Currently, there have been a few studies on the performance analysis of IEEE 802.15.6. Rashwand et al. [5] proposed an analytical model in a saturation condition. However, they investigated a network with nodes having the same user priorities under saturation condition only. In this work, we propose an accurate analytical model for both throughput and energy consumption, and evaluate the performance of the WBAN based on IEEE 802.15.6 protocol with co-located nodes with different user priorities, in an unsaturated condition. We develop an analytical model by using a Discrete-Time Markov Chain (DTMC) to evaluate the performance of the contention-based CSMA/CA protocol. II. IEEE 802.15.6 WBAN S TANDARD The IEEE 802.15.6 standard specifies a communication standard at PHY and MAC layers that should support a variety of medical, consumer electronics (CE) and entertainment applications. The first draft of the IEEE 802.15.6 standard was issued in May 2010 [3]. It defines a MAC layer in support of three different PHY layers. These include Narrowband (NB), Ultra-Wideband (UWB), and Human Body Communications (HBC) layers. At the MAC sub-layer, IEEE 802.15.6 supports two different types of access mechanisms including: contention access and contention-free access. The contention access phase supports either a slotted ALOHA

TABLE I U SER PRIORITIES AND THEIR CW BOUNDS !

"!# "! "#!

$%& "!

$'!

(! (% (!#

(#$

#"!!

&)!!

)!%!!

# !!

*

Fig. 1.

Different frequency bands of IEEE 802.15.6 PHY

based access mechanism or CSMA/CA based access mechanisms. The contention-free access phase supports a scheduled uplink/downlink access scheme as well as an improvised polling/posting based access scheme. In this paper, we focus on performance analysis of HBC PHY layer with CSMA/CA based MAC layer protocol. A. IEEE 802.15.6 HBC PHY specification The IEEE 802.15.6 supports three different PHY layers: Narrow band (NB) PHY, Ultra-Wideband (UWB) PHY, and Human Body Communication (HBC) PHY. The three frequency bands operate at different frequencies and supports different data rates. Figure 1 shows the operating frequencies of these bands in different regions of the world. The IEEE 802.15.6 HBC PHY operates in two frequency bands centered at 16 MHz and 27 MHz with a bandwidth of 4 MHz. Both operating bands are available for the United States, Japan, Korea, and the operating band at 27MHz is abailable for Europe [6]. HBC is the Electrostatic Field Communication (EFC) specification of PHY, which covers the entire protocol for WBAN, such as packet structure, modulation, and preamble/SFD. B. IEEE 802.15.6 CSMA/CA MAC Specification In IEEE 802.1.5.6 standard in a beacon mode with superframe boundaries, a hub divides the time into multiple superframes. Each superframe structure is sub-divided into various access phases. These access phases are basically classified into contention based access, connectionless contentionfree access, and connection-oriented contention-free access. Contention based access is based on either CSMA/CA or slotted Aloha. In CSMA/CA, to obtain a new contended allocation, the node sets its backoff counter as a random in  teger number uniformly distributed over the interval , 

    . The values of and where  depend on the user priorities (UPs), as shown in Table I. When the backoff counter value of a node reaches zero, then the node  ! transmits a packet. The CW value is initially set to . If the transmission is successful, the node receives the expected acknowledgement. IEEE 802.15.6 supports immediate acknowledgment (I-ACK), group acknowledgment (G-ACK), as well as block acknowledgment (B-ACK) [3].  # " If the transmission fails, the value is set to !$ &%'()+*,- for an even number of retries, and is fixed to the previous value for an odd number of retries. The backoff counter value for a node is locked(freezed) if any one of these conditions is true: 1) the channel is busy, 2) the current time is outside the access phases where the node can transmit, 3) the current time is at the start of a CSMA slot within an EAP, RAP, or CAP, but the time between the end

User Priority(k)

.0/2143 5

.0/ 1768

Traffic

0 1 2 3 4 5 6 7

16 16 8 8 4 4 2 1

64 32 32 16 16 8 8 4

Background (BK) Best effort (BE) Excellent effort (EE) Controlled load (CL) Video (VI) Voice (VO) Medical data / network control Emergency / medical report

of the slot and the end of the EAP, RAP, or CAP is not long enough for completing a frame transmission. Moreover, the backoff counter is reset upon decrementing to 0. The backoff counter value is unlocked if: 1) the channel has been idle. 2) The time duration between the current time plus a CSMA slot and the end of the EAP, RAP, or CAP is long enough for completing a frame transmission. Figure 2 shows an example of a CSMA/CA mechanism of the IEEE 802.15.6 as explained above. III. P ERFORMANCE A NALYSIS OF IEEE 802.15.6 WBAN MAC P ROTOCOL The Performance analysis of the CSMA/CA based protocol in WLAN has been a focus of many studies. An analytical model using a discrete-time Markov chain based on CSMA/CA protocol in WLAN was developed by Bianchi, [7]. Later on several researchers proposed enhanced models [10], [11], [12], and [13]. In case of the ZigBee sensor networks, various CSMA/CA based analytical models are proposed based on IEEE 802.15.4 standard [8], [9], [15], [16]. In this paper, we developed an analytical model for evaluating the performance of contention based CSMA/CA procedure of IEEE 802.15.6 under non-saturation conditions. Rashward et al. [5], and Ullah and Kwak [14] investigated the performance analysis of the IEEE 802.15.6. However, to the best of our knowledge, there has been only one article [5] which focused on the CSMA/CA based performance analysis of IEEE 802.15.6 under saturation conditions. A. Discrete-Time Markov Chain Modeling and Analysis In this section, we provide a discrete-time Markov chain model for the analysis of the CSMA/CA based IEEE 802.15.6 MAC. We assume a WBAN with a single hub and nodes with different user priorities in a star topology. We consider a beacon mode with superframe boundaries, as shown in Figure 2. Our analysis focuses on the contention access phases such as EAP, RAP or CAP. In addition, we consider the RTS/CTS mechanism for accessing the medium. The notations used in this analysis are listed in Table II. A node with 9;:$< chooses a backoff counter value from the interval [1,CW], where    CW is the minimum value of the contention window, . The maximum number of backoff stages is bound by a retry limit  . If the number of retries exceeds the predefined retry limit  , the packet is discarded. The CW value is initially set to

$

!

! ! " #

% !

#" &

!

$

!

!

Fig. 2.

&

!

Example of a CSMA/CA mechanism of IEEE 802.15.6 MAC

VARIABLES AND NOTATIONS Notation

Explanation

G? H IJ IK IL

User Priority (UP) Number of nodes of A7B ? Stationary distribution with A7B ? , backoff stage i, backoff counter j Frame transmission probability of A7B ? during a slot time Frame retry limit Frame payload size in bits Successful transmission time Unsuccessful transmission time Probability of occurrence of at least one transmission Probability of successful transmission of a node with A7B ? Contention window size of node with A7B ?

B@MON B K D? / ?ED 3

!& !

!

!

TABLE II

= > ? @ C ?ED D F 3

'

&

we consider an unsaturated condition with packet arrival rate u < of a node with 9;:0< . Hence, we consider a node with an empty state as shown in Figure 3, with queue length of one. The parameters, vw< , xQ< , and yz< represent the collision probability, channel idle probability, and the probability that there is enough time left for frame transmission of 9;:$< , respectively. The vw< , xQ< , and yz< can be expressed as follows: Qac4€ z„ "-{| }{~  }{~r!  v\< < ‚ (2) ƒ