Sage Journals: Discover world-class research

Abstract

In wireless distributed sensor networks, one open problem is how to guarantee the reliable relay selection based on the quality of services diversity. To address this problem, we focus on the reliable adaptive relay selection approach and adaptive QoS supported algorithm,

based on which we present a Markov chain model, in consideration of different packet states and error control algorithm assignment. The mathematical analyses and NS-2 simulation results show that the proposed relay selection approach could perform better in terms of saturation throughput, reliability, and energy efficiency, compared with the traditional approaches. More importantly, the quality of real-time multimedia streaming is improved significantly, in terms of decodable frame ratio and delay.

1. Introduction

It is well known that delivering and transmitting various data over wireless distributed and dynamic networks is a very challenging task. Wireless communication link is a dynamic environment. Data transmission over wireless channel suffers from the limited bandwidth, high packet loss rate, dynamic changes of channel state, sensor nodes mobility, channel competition, and so forth [1]. These constraints and challenges, in combination with the delay and loss of sensitive nature of multimedia applications [2], make QoS provisioning over wireless sensor networks a challenging proposition.

The data delivery of different applications inherently has different QoS requirements [3, 4] such as throughput, real time performance, playable frame rate, reliability and energy efficiency, and so forth. There is an open problem in the relay selection approach design: how to take advantage of the requirements information of QoS diversity in the wireless distributed sensor networks to find out the optimal relay for forwarding data packets from the potential relays.

In this paper, we investigate how to select the optimal relay nodes to provide the reliable and effective QoS provision and satisfy the diversity requirements of application services.

The rest of the paper is organized as follows. The related work and our work are elaborated in Section 2. Section 3 describes the system models and problem description. In Section 4, we design a Markov chain model to describe process of data communication in wireless distributed sensor network and give the theoretical analysis model to show the achieved performance. The supported scheme of QoS diversity is used to realize reliable, efficient, and robust communication in wireless distributed sensor network, which is shown in Section 5. Section 6 proposes the QoS supported adaptive relay selection (ECA-Q-RS) approach, followed by the details of implementation. Simulation and mathematics results are given in Section 7. Finally, we conclude the paper in Section 8.

2. Related Work

Recent research shows that the traditional cooperative strategies, such as amplify-and-forward (AF), decode-and-forward (DF), or estimate-and-forward (EF), work well in networks of static channel state information (CSI) but have scalability limitations that degrade performance in larger or rapidly moving networks. So, research on AF and DF has drawn researchers’ attention which focuses on outdated channel estimates [5], outdated CSI [6], outage probability and bit error probability [7], and mobile relays [8], in order to reduce the complexity in management about mobile sensor nodes and simplify some essential procedures such as reliability guarantee and wireless resource allocation. In additon, error control scheme is usually used to improve the reliability for wireless distributed sensor networks, which can be realized by automatic repeat request (ARQ) [9], forward error correction (FEC) [10], or hybrid ARQ (HARQ) [11, 12].

How to combine error control protocol and cooperative technology in wireless distributed sensor networks has been fairly studied in the literature [13–15]. Lo et al. [13] present a decentralized relay selection protocol, which supports HARQ transmission where relay nodes forward parity information to the destination node in the event of a decoding error. A high energy efficiency adaptive cooperative error control mechanism was proposed in [14], which predicts frame loss rate using the model of GM (1,1) and adjusts the FEC parameter according to the energy efficiency of the sensor nodes. In addition, Wang et al. [15] designed a deterministic linear network coding scheme for reliable communication against multiple failures in wireless sensor networks. Moreover, the energy of sensors depends on the battery mainly, which is limited and difficult to replace or recharge. Therefore, a data forwarding scheme termed dynamic energy-based relaying proposed in [16] tries to forward data packets toward these sensor nodes which are closer to the data sink with optimal distance.

Besides, QoS requirements [17] play a major part in relay selection mechanism for real-time multimedia applications since they can be applied to find relay nodes under application's diversity requirements.

As we know, the static scheme has several drawbacks. That means that the data packets will be treated with the same way, which may introduce a great inefficiency due to the characteristics of ARQ, FEC, and HARQ. Hence, an adaptive error control mechanism was proposed in [18], which can improve the performance of wireless sensor network by scheduling the optimal error control protocols according to the feature of distance between sending node and next-hop receiving node with the energy efficiency.

Existing relay selection protocols do not have explicit consideration of cross-layer design and interactions, as well as combination of adaptive error control schemes and QoS diversity guarantees for different application services. This paper proposes an efficient and robust QoS supported adaptive relay selection protocol (ECA-Q-RS) for service diversity in wireless distributed sensor networks. The purpose is to explore a relay selection protocol with satisfying QoS diversity. The proposed ECA-Q-RS introduces a series of improved error control mechanisms. First, we present a Markov chain model based on three different states of the wireless data packets and two different error control algorithms. Second, a reliability aware QoS analysis model to detect the performance is suggested. The aim is to estimate whether we should update the error control scheme by the threshold of distance or signal-to-noise ratio (SNR) or not. Finally, considering the characteristics of QoS diversity requirement, we introduce an adaptive QoS supported strategy into relay selection protocol, based on cross-layer design, to alleviate the congestion and optimize the limited resource allocation.

3. System Model

In this section we present the wireless distributed sensor network model which is composed of Mica2 sensor nodes, which are comprised of ATmega128L processor and CC1000 radio module. In order to analyze the variation characteristics of SNR between hops, the log-distance path loss model is used, which is able to approximate the effects of the signal propagating through the wireless channel. In this model, d denotes the distance between the sending node and receiving node. The received power of transmitter node with d is given by

\begin{matrix} P_{r} (d) = P_{t} - P (d_{0}) - 10 β l g (\frac{d}{d_{0}}) . \end{matrix}

(1)

Here, let β denote the path loss exponent, and assume that it is set to 3. The close-in reference distance $d_{0}$ is determined by the physical attribution of the sensor node. And the SNR at the receiver node can be calculated as

\begin{matrix} SNR = P_{r} (d) - P_{n}, \end{matrix}

(2)

where $P_{n}$ is the noise power and Mica2 node is implemented with no coherent FSK modulation scheme. So the bit error rate of the wireless channel is given by

\begin{matrix} P_{b} = \frac{1}{2} e^{SNR (B_{N} / 2 R_{radio})}, \end{matrix}

(3)

where $B_{N}$ is the noise bandwidth and the data rate of CC1000 is set to $R_{radio}$ .

In this network model, HARQ or ARQ scheme is applied in sending nodes, receiving nodes, and relay nodes. Let $P_{ARQ}$ denote the packet error rate with ARQ, which could be derived as follows

\begin{matrix} P_{ARQ} = 1 - {(1 - P_{b})}^{l_{DATA}}, \end{matrix}

(4)

where the sum of data link layer frame header length and frame check sequence size is α and the packet length of ACK is $l_{ACK}$ , so the data packet size $l_{DATA}$ can be given as α plus $l_{ACK}$ . Here, the value of $l_{DATA}$ and $l_{ACK}$ could be obtained according to the MAC protocol in sensor node.

Furthermore, let $P_{HARQ}$ denote the packet error rate with HARQ, which could be obtained by

\begin{matrix} P_{HARQ} = 1 - {(\sum_{l = K}^{M} (\begin{pmatrix} M \\ l \end{pmatrix}) {(1 - P_{b})}^{l} {P_{b}}^{M - l})}^{l_{DATA}} . \end{matrix}

(5)

4. Reliability Aware QoS Analysis Model

4.1. Markov Chain Model

When the packet has been sent into the wireless channel, it may be in three states: (1) received state, (2) drop state, or (3) aware state. Let $P_{EC}$ denote the packet error rate, and let $N_{\max}$ be the maximum retransmission time. N is called retransmission time. Any packet would be in the aware state if it has been sent by the sender node. In particular, there are two error control mechanisms, HARQ and ARQ, in our work. When HARQ is used, N is set to $N_{HARQ}$ . When ARQ is used, N is set to $N_{ARQ}$ . Because the stochastic processes of outage of wireless channel and packet error are independent each other, the future outage of wireless link has nothing to do with the past damaged packets. Hence, it is able to model the process with the discrete-time Markov chain depicted in Figure 1.

Figure 1

Markov chain model for the error control scheme.

In this Markov chain, the nonnull one-step transition probabilities from aware state into received state are listed as follows:

\begin{array}{l} P {(Aware state, N) ⟶ Received state} = 1 - P_{EC}, \\ N = 0, \end{array}

(6)

\begin{array}{l} P {(Aware state, N) ⟶ Received state} = 1 - P_{EC}^{N}, \\ 0 < N \leq N_{\max} . \end{array}

(7)

The probability of packet retransmitted by the relay nodes is given by following formula:

\begin{array}{l} P {(Aware state, N) ⟶ (Aware state, N + 1)} = P_{EC}^{N}, \\ 0 < N \leq N_{\max} . \end{array}

(8)

For the transition probabilities from other states into drop state, we have

\begin{array}{l} P {(Aware state, N_{\max}) ⟶ Drop state} = P_{EC}^{N_{\max} + 1}, \\ 0 < N \leq N_{\max} . \end{array}

(9)

When the physical characteristics, distance, and the value of $N_{\max}$ are all known, we can calculate the transition probabilities between three states, which are unique. Therefore, the analytical model of saturation throughput, packet dropping rate, average delay, and energy efficiency would be given and discussed in Section 4.2.

4.2. QoS Analysis Model

Let $S_{EC}$ denote the normalized saturation throughput, which is the ratio of the payload length to all the lengths of data packets after these packets have been retransmitted at the N time successfully. So, we have

\begin{array}{l} S_{EC} = \frac{l_{payload}}{l_{DATA}} ((1 - P_{EC}) + P_{EC} (1 - P_{EC}) \\ + P_{EC}^{2} (1 - P_{EC}) + \dots + P_{EC}^{N_{\max}} (1 - P_{EC})) \\ = \frac{l_{payload}}{l_{DATA}} (1 - P_{EC}^{N_{\max} + 1}) . \end{array}

(10)

Here, $l_{payload}$ is the length of payload. As shown in Figure 1, the data packet would be dropped only when N is larger than $N_{\max}$ . Consequently, the packet dropping rate after the error control process could be achieved by the following equation:

\begin{matrix} P_{EC} = P_{EC}^{N_{\max} + 1} . \end{matrix}

(11)

The average retransmission time $N_{avg-EC}$ can hence be given as

\begin{matrix} N_{avg-EC} = P_{EC} + P_{EC}^{2} + \dots + P_{EC}^{N_{\max}} . \end{matrix}

(12)

According to (11), we can obtain the solution of average delay, which is shown as (12):

\begin{matrix} T_{ARQ} = T \frac{1 - P_{EC}^{N_{\max} + 1}}{1 - P_{EC}} . \end{matrix}

(13)

Here, let T denote the round trip time of one packet. The energy efficiency η is defined as (14), which is a suitable metric that could capture the energy and reliability constraints:

\begin{matrix} η = \frac{E_{effi}}{E_{total}} (1 - P) . \end{matrix}

(14)

η means the ratio of the energy of payload $E_{effi}$ to the total energy consumption $E_{total}$ . Therefore, the energy efficiency of error control scheme $η_{EC}$ is given as

\begin{matrix} η_{EC} = \frac{E_{effi}^{ARQ}}{E_{total}^{ARQ}} (1 - P_{EC}) = \frac{l_{payload}}{l_{DATA}} (1 - P_{EC}) . \end{matrix}

(15)

4.3. Results and Discussion

In our previous research achievements [18], the performance of ARQ or HARQ scheme has a close relationship with distance and $d_{0}$ at first. Then, the variation characteristic of ARQ or HARQ scheme performance is presented by mathematical analyses in different SNR. All the above evaluations are based on the equations from (10) to (15).

On the one hand, Figure 2 shows four performance metrics as a function of distance. We only consider two different $d_{0}$ , which are 15 m and 25 m. As illustrated in Figure 2, because the channel state is perfect and the frame error rate is extremely small, the performance of ARQ scheme is maintained relatively well with the increasing of the distance. However, the saturation throughput and energy efficiency of ARQ decrease to zero rapidly. Meanwhile, the packet dropping rate and average delay of this scheme increase quickly when the distance between sender node and next-hop receiver node is greater than 40 m, and $d_{0}$ is 15 m. Hence, the distance between sender node and next-hop receiver node has a constant value dependent on variation of $d_{0}$ , which is 40 m for ARQ scheme. It is clearly noticed that there is a distance threshold value of ARQ scheme, which is 40 m, when $d_{0}$ is 15 m. Likewise, the distance threshold value is 70 m, 30 m, and 60 m, respectively, for ARQ ( $d_{0}$ is 25 m), HARQ ( $d_{0}$ is 15 m), and HARQ ( $d_{0}$ is 25 m). Above all, there is always constant distance threshold value of different error control mechanisms with different $d_{0}$ . Consequently, we can obtain the number of relay nodes, and select the best relay node according to this result.

Figure 2

Performance analyses of ARQ and HARQ with different distances and $d_{0}$ .

On the other hand, the change trend of four performance metrics as a function of SNR is shown in Figure 3. It was obvious that the performance of ARQ is maintained gradually perfect with increasing of SNR. Specially, the saturation throughput and energy efficiency of ARQ increase to 1 rapidly; meanwhile, the packet dropping rate and average delay of this scheme decrease fast when the SNR between sender node and next-hop receiver node is greater than 16 dB. Hence, SNR between sender node and next-hop receiver node has a constant value, which is 16 dB for this scheme. There is apparently a SNR threshold value of ARQ scheme. Likewise, the SNR threshold value is 20 dB for HARQ. Therefore, different error control mechanisms have always one constant SNR threshold value. We can choose the best relay node as the next-hop receiving node according to this conclusion.

Figure 3

Performance analyses of ARQ and HARQ with SNR.

5. Diversity of QoS Supported Scheme

In this section, we define six error control schemes: (1) ARQ with distance ( $d_{0} = 15$ ), (2) ARQ with distance ( $d_{0} = 25$ ), (3) HARQ with distance ( $d_{0} = 15$ ), (4) HARQ with distance ( $d_{0} = 25$ ), (5) ARQ with SNR, and (6) HARQ with SNR. On the basis of the above QoS analysis model based on error control, the variation trend of saturation throughput, packet dropping rate, average delay, and energy efficiency is illustrated in Figure 4.

Figure 4

Analytical research of four performance metrics between six different error control schemes.

From Figure 4, we discover the relationship of QoS supported capacity with the above error control scheme, which is listed as follows.

(1)

Energy efficiency: ARQ with SNR > HARQ with SNR > ARQ with distance ( $d_{0} = 25$ ) > HARQ with distance ( $d_{0} = 25$ ) > ARQ with distance ( $d_{0} = 15$ ) > HARQ with distance ( $d_{0} = 15$ ).

(2)

Average delay: HARQ with distance ( $d_{0} = 15$ ) < ARQ with distance ( $d_{0} = 15$ ) < ARQ with distance ( $d_{0} = 25$ ) < HARQ with distance ( $d_{0} = 25$ ) < HARQ with SNR < ARQ with SNR.

(3)

Packet dropping rate: ARQ with SNR > HARQ with SNR > ARQ with distance ( $d_{0} = 25$ ) > HARQ with distance ( $d_{0} = 25$ ) > ARQ with distance ( $d_{0} = 15$ ) > HARQ with distance ( $d_{0} = 15$ ).

(4)

Saturation throughput: ARQ with SNR > HARQ with SNR > ARQ with distance ( $d_{0} = 25$ ) > HARQ with distance ( $d_{0} = 25$ ) > ARQ with distance ( $d_{0} = 15$ ) > HARQ with distance ( $d_{0} = 15$ ).

We found out that the trend of saturation throughput, packet dropping rate, and energy efficiency is quite opposite to the one of average delay. Therefore, the application service could select the optimal QoS supported scheme to satisfy the diversity requirement from Table 1. Here, QoS scheme I is composed of the guarantee ability of saturation throughput, packet dropping rate, and energy efficiency. In addition, QoS scheme II is comprised of real-time guarantee ability. Particularly, the more the digit of the error control scheme in Table 1, the better its guarantee capacity.

Table 1

QoS supported scheme.

Error control scheme	Scheme I	Scheme II
ARQ with SNR	6	1
HARQ with SNR	5	2
ARQ with distance $(d_{0} = 25)$	4	3
HARQ with distance ( $d_{0} = 25)$	3	4
ARQ with distance $(d_{0} = 15)$	2	5
HARQ with distance $(d_{0} = 15)$	1	6

6. High Reliable Relay Selection Approach

6.1. Relay Selection Algorithm

If we can obtain the end to end distance d, the optimal relay node should be selected according to distance threshold, and also the number of relay nodes is given, which are as shown in the following equations:

\begin{matrix} d = {\begin{cases} D_{TE}, & d < D_{TE}, H_{\max} = 1, \\ H_{\max} D_{TE}, & d \mod D_{TE} = 0, H_{\max} = \frac{d}{D_{TE}}, \\ H_{\max} D_{TE}, & d \mod D_{TE} \neq 0, H_{\max} = [\frac{d}{D_{TE}}] + 1, \end{cases} \end{matrix}

(16)

where $D_{TE}$ is the distance threshold value; let mod denote the calculation of solving the modulus operator. [ $d / D_{TE}$ ] indicates the rounding calculation. That means that the integer part is intercepted and the decimal fraction is dropped. The cooperative hops $H_{\max}$ can be obtained if d is known; the number of relay nodes is $H_{\max}$ minus 1 simultaneously.

In a word, the work flow of relay selection based on distance is as follows.

(1)

To obtain the value of $d_{0}$ with sensor node and d according to the actual measurement.

(2)

Calculate $D_{TE}$ according to the error control aware analytical model.

(3)

Get $H_{\max}$ and the number of relay nodes on the basis of (16).

(4)

If the distance is less than or equal to $D_{TE}$ , it would be selected as the relay node.

If we can obtain the SNR of wireless channel, the optimal relay node should be chosen according to SNR threshold. The process of relay selection based on SNR is as follows.

(1)

The SNR of the wireless channel is acquired with real-time detection.

(2)

Acquire the SNR threshold SNR_TE based on the QoS analytical model.

(3)

If the SNR of one sensor node is less than or equal to SNR_TE, it would be selected as the relay node.

Particularly, relay selection based on SNR should be used first when the wireless channel is stable and perfect; otherwise, relay selection based on distance has to be employed promptly.

6.2. Implementation of Adaptive Relay Selection

In this subsection, the reliability aware and diversity of QoS supported adaptive relay selection mechanism (ECA-Q-RS) is proposed, which is implemented in wireless distributed sensor network. Because the relay node is selected based on characteristic of error control in ECA-Q-RS, the performance of the proposed mechanism can be evaluated by the following formula:

\begin{matrix} Q_{ECA-Q-RS} = \sum_{t = 1}^{H_{\max}} Q_{C} (t) + q, \end{matrix}

(17)

where $Q_{ECA-Q-RS}$ denotes the performance metrics of ECA-Q-RS, which include saturation throughput, packet dropping rate, average delay, and energy efficiency. Let $Q_{C} (t)$ denote the above four performance metrics when d is larger than $D_{TE}$ or SNR is greater than SNR_TE. q is used to record these metrics when d is less than $D_{TE}$ or SNR is less than SNR_TE.

Afterwards, we present the basic idea of the ECA-Q-RS and its implementation at sender node, receiver node, and relay nodes in detail, which is illustrated as follows.

Sender Node

(1)

Carry out the QoS supported scheme. The guarantee priority of saturation throughput, packet error rate, and energy efficiency, or average delay is appointed according to the requirement of application service.

(2)

At the data link layer, the optimal error control and QoS supported scheme are chosen on the basis of Table 1. Moreover, the value of $D_{TE}$ and $H_{\max}$ is obtained based on the QoS analytical model.

(3)

Relay selection mechanism based on distance is implemented when d and $d_{0}$ are known, or the wireless channel is unstable and poor. Relay selection mechanism based on SNR is implemented when channel state information is known, or the wireless link is perfect.

(4)

Starting to send data packets. If ACK packet is received, new data packets are sent, and retransmission time-out is used for each packet at the same time.

(5)

When the timer matures, or NACK packet is received, the optimal error control scheme is started according to the QoS scheme.

Relay Node

(6)

Select the optimal relay node from candidate nodes based on distance or SNR threshold.

(7)

Steps (3), (4), and (5) are implemented repeatedly in proper order until the data packet is received successfully or discarded actively.

Receiver Node

(8)

If ARQ scheme is used, checksum is calculated and tested. If HARQ scheme is used, checksum testing and FEC encoding are implemented.

(9)

If the result obtained is right, the data packet is accepted, and ACK packet is sent simultaneously; otherwise, it is rejected, and NACK packet is sent at the same time.

(10)

Deliver the correct data packet to the upper layer.

7. Performance Evaluation

In this work, we use NS-2 and VC++6.0 to simulate, analyze and evaluate the performance of ordinary data transmission and multimedia streaming using ECA-Q-RS, compared with max SNR relay selection and distance aware relay selection through two group experiments. The experimental data is the average value after 100-time simulation and mathematical analysis.

7.1. Simulation Parameters

In experiment 1, 50 sensor nodes move in an 800 m × 800 m rectangular region. α is 11 bytes. $l_{ACK}$ is set to 7 bytes. Noise bandwidth $B_{N}$ is 30 kHz. The data rate of CC1000 $R_{radio}$ is equal to 38.4 kBaud. Round trip time of one frame transmission at data link layer T is 0.01 second. The mobility model is the random waypoint model. Each sensor node moves towards the destination node at a speed uniformly chosen between the minimal speed 0 m/s and maximal speed 10 m/s. The MAC protocol is IEEE 802.15.4 protocol. We change $d_{0}$ from 15 m to 30 m to investigate the impact of node physical attribution on performance.

For multimedia traffic of experiment 2, we use a medium quality MPEG4 video clip from the movie forman_qcif.yuv, which consists of 400 video frames. The structure of the group of pictures is IBBPBBPBBPBB ( $N = 12$ , $M = 3$ ). The video frame rate is 25 frames per second. The maximum transmission unit is 1024 bytes.

7.2. Simulation Results

Two case studies are designed and conducted, with the variation of bit error rate $P_{b}$ and the simulation time, respectively. The main metrics for performance evaluation of the ECA-Q-RS for supporting the application services of QoS diversity are listed as follows.

(1)

Decodable frames rate: it considers the dependency between different MPEG4 video frames.

(2)

Average delay: we only calculate the end to end delay of the decodable frames.

(3)

Throughput: the total size of data packets received successfully by the receiver node.

(4)

Energy efficiency: the number of obtained decodable video frames per unit of energy consumption.

7.2.1. Experimental Results with Channel State

Figure 5 shows three performance metrics as a function of $P_{b}$ in experiment 1. The result indicates that channel state has a significant impact on quality of data transmission in wireless sensor networks. We can observe tremendous improvement of performance using ECA-Q-RS, as compared with other two mechanisms. Even at a more poor state of the wireless channel, the quality using ECA-Q-RS is maintained in a good condition, while the saturation throughput and energy efficiency using max SNR relay selection and distance aware relay selection reduce and become close to zero gradually. The reasons are that not only the supported mechanism of QoS diversity minimizes the average delay but also the adaptive error control strategy improves the energy efficiency; as a result, the network throughput utilization increased.

Figure 5

Performance with varying channel state ( $P_{b}$ ).

7.2.2. Experimental Results with QoS Supported Capacity

Figure 6 shows the five performance metrics as a function of different relay selection schemes in experiment 2.

Figure 6

Performance with multimedia communication.

Figure 6(a) shows the result of the decodable frame rate. As the transmission rate of multimedia data increases, the collision probability of data packets transmission increases significantly, leading to an unstable and dynamic decreasing tendency of decodable frame rate. The result demonstrates tremendous improvement of decodable frames rate with ECA-Q-RS as compared with other mechanisms. On this basis, we determine that ECA-Q-RS can prime accommodate the poor and dynamic wireless network environment. This obvious improvement depends on their stable relay selection scheme based on QoS diversity supported mechanism and adaptive error control strategy.

The real-time performance is shown in Figure 6(b). The average delay using ECA-Q-RS achieves a tremendous improvement. The main reason is that the error control aware adaptive relay selection strategy enhances the bandwidth and frequency spectrum utilization, as well as propagation path. As shown in Figure 6(c), the ECA-Q-RS always performs better than the other two mechanisms. The throughput gain we have is in the improvement of hop-by-hop throughput and reduction in bandwidth consumption.

Figure 6(d) provides packet error rate for each video frame in different mechanisms. When using ECA-Q-RS, the packet error rate fluctuates a little, and the average is the lowest. These adaptive schemes in ECA-Q-RS are not only able to achieve higher decodable frame rate and network throughput but also improve the stability of the video transmission. Figure 6(e) shows the energy utilization efficiency of different mechanisms. The enhancement of data exchange gain and adaptive error control can greatly reduce energy consumption for data transmission. Comparing with the other two schemes, ECA-Q-RS achieves significant improvement in energy utilization efficiency.

8. Conclusions and Future Work

There is a great potential for relay selection and error control to enhance the performance of QoS diversity services and resource utilization. The purpose of this work is to overcome all kinds of limitations; we propose a reliability aware and QoS supported relay selection (ECA-Q-RS).

The main contributions in our work are as follows. First, considering the characteristics of different error control schemes, we introduce the Markov chain model based on ARQ and HARQ to study the characteristics of QoS. Second, a QoS supported strategy with dynamic priority is designed to ensure that packets have a better chance to be forwarded. Finally, we present adaptive relay selection algorithm, working together with the above QoS supported mechanism, in order to reduce the ratio of potential damaged or lost opportunities.

The mathematics and simulation results demonstrate that, compared with the existing typical relay selection mechanisms, ECA-Q-RS greatly improves the data transmission quality and achieves significant gains in terms of throughput and energy efficiency. As a result, we determine that the proposed mechanism is feasible for data diversity communication in wireless distributed sensor network.

Our future work focuses on the implementation and validation of the proposed adaptive relay selection algorithms on the prototype system. Besides, one possible future direction is to study the channel prediction approach and the tradeoff between performance and complexity.

Footnotes

Conflict of Interests

The authers declare that they have no financial and personal relationships with other people or organizations that can inappropriately influence their work; there is no professional or other personal interest of any nature or kind in any product, service, and/or company that could be construed as influencing the position presented in, or the review of the paper.

References

Xia

QoS challenges and opportunities in wireless sensor/actuator networks

Sensors 2008 8 2 1099 1110

2-s2.0-40849094798

Bai

G. W.

Oladosu

Williamson

Performance benchmarking of wireless web servers

Ad Hoc Networks 2007 5 3 392 412

2-s2.0-33845894741

10.1016/j.adhoc.2006.01.001

Bai

G. W.

Tao

J. J.

Shen

A link-lifetime based dynamic source routing protocol (LTDSR) for multimedia over MANETs

Journal of the Chinese Institute of Engineers 2010 33 5 761 768

2-s2.0-77958090684

Shen

Bai

G. W.

Zhao

Tang

Z. M.

An adaptive opportunistic network coding mechanism in wireless multimedia sensor networks

International Journal of Distributed Sensor Networks 2012 2012 13

565604

10.1155/2012/565604

Michalopoulos

D. S.

Suraweera

H. A.

Karagiannidis

G. K.

Amplify-and-forward relay selection with outdated channel estimates

IEEE Transactions on Communications 2012 60 5 1278 1290

Y. B.

On the design of relay selection strategies in regenerative cooperative networks with outdated CSI

IEEE Transactions on Wireless Communications 2011 10 9 3086 3097

2-s2.0-80052939367

10.1109/TWC.2011.072511.110077

Bao

V. N. Q.

Kong

H. Y.

Performance analysis of multi-hop decode-and-forward relaying with selection combining

Journal of Communications and Networks 2010 12 6 616 623

2-s2.0-79251625776

Miranda

Natalizio

Razafindralambo

Adaptive deployment scheme for mobile relays in substitution networks

International Journal of Distributed Sensor Networks 2012 2012 9

128904

10.1155/2012/128904

Fairhurst

Wood

Advice to link designers on link Automatic Repeat reQuest (ARQ)

RFC 3366, IETF, August 2002

10.

Watson

Luby

Vicisano

Forward Error Correction (FEC) Building Block

RFC 5052, IETF, August 2007

11.

Chase

Code Combining: a maximum-likelihood decoding approach for combining an arbitrary number of noisy packets

IEEE Transactions on Communications 1985 33 5 385 393

2-s2.0-0022054524

12.

Stanojev

Simeone

Bar-Ness

Kim

D. H.

Energy efficiency of non-collaborative and collaborative Hybrid-ARQ protocols

IEEE Transactions on Wireless Communications 2009 8 1 326 335

2-s2.0-61349155628

10.1109/T-WC.2009.071403

13.

C. K.

Heath

R. W.

Vishwanath

The impact of channel feedback on opportunistic relay selection for hybrid-ARQ in wireless networks

IEEE Transactions on Vehicular Technology 2009 58 3 1255 1268

2-s2.0-64049112628

10.1109/TVT.2008.928896

14.

Jin

Bai

G. W.

A cooperative FEC based on the model of GM (1, 1) and IPv6 for wireless multimedia sensor networks

Journal of Convergence Information Technology 2012 7 18 230 239

15.

Wang

X. M.

Zhang

S. K.

An efficient reliable communication scheme in wireless sensor networks using linear network coding

International Journal of Distributed Sensor Networks 2012 2012 11

605494

10.1155/2012/605494

16.

Deng

C.-H.Huang

Scott

Han

Y. S.

An online relay selection scheme in power controllable wireless sensor networks

International Journal of Distributed Sensor Networks 2012 2012 13

213598

10.1155/2012/213598

17.

Cho

S.-R.

Wan

Huang

K. B.

QoS provisioning relay selection in random relay networks

IEEE Transactions on Vehicular Technology 2011 60 6 2680 2689

2-s2.0-79960363457

10.1109/TVT.2011.2153220

18.

Jin

D.-G.

Bai

G.-W.

Chang

J.-Y.

Adaptive error control scheme for wireless sensor networks based on hops and communication distance

Control Theory and Applications 2011 28 4 596 600

2-s2.0-79958754655

High Reliable Relay Selection Approach for QoS Provisioning in Wireless Distributed Sensor Networks

Abstract

1. Introduction

2. Related Work

3. System Model

4. Reliability Aware QoS Analysis Model

4.1. Markov Chain Model

4.2. QoS Analysis Model

4.3. Results and Discussion

5. Diversity of QoS Supported Scheme

6. High Reliable Relay Selection Approach

6.1. Relay Selection Algorithm

6.2. Implementation of Adaptive Relay Selection

7. Performance Evaluation

7.1. Simulation Parameters

7.2. Simulation Results

7.2.1. Experimental Results with Channel State

7.2.2. Experimental Results with QoS Supported Capacity

8. Conclusions and Future Work

Footnotes

Conflict of Interests

References