Design and Evaluation of a Two-Channel Slotted Sense Multiple Access Protocol for Dynamic Wireless Sensor Networks|RISS 상세보기

다국어 초록 (Multilingual Abstract)

To design an efficient Media Access Control (MAC) protocol for supporting the time-sensitive applications with a Wireless Sensor Network (WSN), the data packet collector or sink node has to define Data Acquisition Cycle Time (DACT), which is the time bound for collecting the data packets from all the deployed nodes in WSNs. Therefore, it is a very important design objective to shorten the DACT period since a small DACT allows many time-critical applications with varying time bound requirements for collecting the data packets from WSNs. The subject is highly relevant to the time-sensitive industrial applications in which a WSN is used as a monitoring and control application to continuously monitor the working people in industry fields, and give the essential protection against any mishaps. However, the lowering of DACT is a complex issue in practice when the relevant industry application demands for a high data transmission efficiency from dynamic WSNs.

For improving a DACT, we generally employ two popular approaches: one is time slot reuse technique and another is sharable time slot technique. Contemporary slot reuse techniques often suffer from the irregular signal interference problems since the transmission range of a node always falls short than the interference range in WSNs. Therefore, a systematic design rule is required so that the advantages we gain from the slot reuse techniques do not introduce any unnecessary communication hurdles. Meanwhile, the use of a sharable time slot allows an opportunistic parallel transmission of data packets to improve a DACT. Obviously, the data transmission efficiency could be accelerated if we were able to utilize both the slot reuse and sharable slot techniques together. However, the sharable slot technique requires allocating an exponentially increasing size of time slots as we move toward a sink node from leaf nodes. The usage of unequal sharable slots for different tree levels makes the slot reuse technique a difficult choice. Therefore, it is quite challenging to design a sensor MAC protocol that takes the benefits from both sharable time slots and slot reuse techniques.

We, therefore, suggest a data transmission model that uses an equal size sharable time slots by utilizing data aggregation and filtering techniques with two frequency channels. In the proposed model, each sensor node allocates the essential number of sharable time slots depending on the tree level in which it belongs to. Additionally, the sensor nodes can allocate non-interfering frequency channels distributedly using simple divide and modular functions. Finally, all sensor nodes schedule their data transmission time according to their tree level, and transmit their data packets toward the sink node at an alternate sharable slot. Within a sharable slot, the proposed approach allows an opportunistic delivery of data packets i.e., two nodes at the same tree level can transmit their data packets in parallel if their transmissions do not violate the contention rules. Moreover, it allows a channel-assisted data transmission in which the sensor nodes at an alternate tree level can use a couple of frequency channels for transmitting data packets, and a spatial data transmission in which the four-hop away sensor nodes along the vertical direction can use the same frequency channel for transmitting data packets. Thus, our proposed MAC protocol is capable of utilizing several slot reusing techniques in a compound way.

The analytical study presented in this thesis work demonstrates that our proposed sensor MAC protocol is capable of reducing approximately 60% of DACT compared to other recent sensor MAC protocols. In the simulation experiments, it is shown that the proposed protocol is able to achieve a remarkably good performance compared to others in terms of Packet Delivery Ratio (PDR), and the energy consumption and balancing. In particular, it just outperforms other competing MAC protocols in improving PDR efficiency for rectangular dimensional areas. Last but not least, all the above gains that we have achieved from our approach by using two orthogonal frequency channel only.

번역하기

목차 (Table of Contents)

Acknowledgment ……………………………………………………………………..........v
Abstract …………………………………………………………………………………. vi
Table of Contents ……………………………………………………………………….......viii
List of Figures …………………………………………………………………………..........xi
List of Tables …………………………………………………………………………….. xii

Acknowledgment ……………………………………………………………………..........v
Abstract …………………………………………………………………………………. vi
Table of Contents ……………………………………………………………………….......viii
List of Figures …………………………………………………………………………..........xi
List of Tables …………………………………………………………………………….. xii
List of Acronyms ………………………………………………………………………........xiii
Chapter 1. Introduction ……………………………………………………………….......1
1.1. Dynamic Wireless Sensor Networks ……………………………………………. 2
1.2. Issues and Challenges …………………………………………………………… 4
1.3. Previous Works …………………………………………………………………. 7
1.4. Our Approach …………………………………………………………………… 10
1.5. Evaluation Methods ……………………………………………………………... 12
1.6. Thesis Organization ……………………………………………………………... 12
Chapter 2. Background ………………………………………………………………... 14
2.1. Network Model ………………………………………………………………….. 15
2.2. Problem Identifications …………………………..……………………………... 16
2.3. An Overview of Proposed Approach ………………………………………........21
2.4. Notations and Definitions ……………………………………………………….. 28
Chapter 3. A Two-Channel Slotted Sense Multiple Access Protocol ...…………. 30
3.1. Protocol Structure ……………………………………………………………….. 31
3.2. Tree Construction and Maintenance ……………………………………………..32
3.2.1. Identification of Bi-Directionally Reliable Link …………………………….32
3.2.2. Reliable Tree Construction and Maintenance …………………………….33
3.3. Time Synchronization …………………………………………………………… 34
3.4. Slot and Channel Allocation scheme …………………………………………….36
3.4.1. Data Transmission Model ………………………………………………… 36
3.4.2. Estimating the Maximum Number of Aggregated Nodes ……………..39
3.4.3. Estimating the Theoretical Size of a Sharable Slot ………………………41
3.4.3.1. Determining the Transmission Time of a MAC Payload ……………… 41
3.4.3.2. Calculating the size of a Sharable Slot ………………………………... 42
3.4.4. Slot Allocation Functions ………………………………………………… 43
3.4.5. Channel Allocation Functions ……………………………………………. 46
3.4.5.1. The use of channel-assisted slot reuse technique ……………………… 46
3.4.5.2. The use of spatial slot reuse technique ………………………………… 48
3.5. Data Transmission Scheduling ………………………………………………….. 50
3.6. Reliable Data Transmission ……………………………………………………... 52
3.7. Data Aggregation and Filtering …………………………………………………. 54
3.7.1. Data Aggregation …………………………………………………………. 54
3.7.2. Data Filtering ……………………………………………………………... 55
Chapter 4. Protocol Evaluation ……………………………………………………….. 56
4.1. An Analytical Study of DACT ……….…………………………………………. 57
4.1.1. Comparing DACT of competing protocols …………………………….... 57
4.1.2. Lower bound of sSlotLen ………….……………………………………… 58
4.1.2.1. Estimating the depth and neighbor nodes of a tree……….................60
4.1.2.2. Determining the size of a sharable slot ……………………………….. 62
4.1.3. Lower bound of DACT for different slotted protocols …………………64
4.2. Performance Evaluation …………………………………………………………. 65
4.2.1. Simulation Setup ………………………………………………………….. 65
4.2.2. Evaluation Scenarios ……………………………………………………... 67
4.2.3. Determination of Data Acquisition Cycle Time …………………………..68
4.2.4. Evaluation with Scenario I ……………………………………………….. 70
4.2.4.1. Packet Delivery Ratio for Scenario I ……………………………….…. 71
4.2.4.2. Energy Consumption for Scenario I .…………………………………... 73
4.2.5. Evaluation with Scenario II ……………………………………………… 74
4.2.5.1. Packet Delivery Ratio for Scenario II ………………………………….. 75
4.2.5.2. Energy Consumption for Scenario II ….……………………………….. 77
Chapter 5. Conclusions and Future Works ………………………………………….. 79
5.1. Conclusions ………………………………………………………………………. 80
5.2. Future Works …………………………………………………………………….. 81
Bibliography ……………………………………………………………………………. 82
Vita ……………………………………………………………………………………….. 85
List of Publications …………………………………………………………………….. 86

참고문헌 (Reference)

1. UHF Fading in Factories, Mcgillem, C. D., Rappaport, T. S., 7(1), pp. 40-48, , 1989

2. Z-MAC : A Hybrid MAC for Wireless Sensor Networks, Rhee , I. ; Warrier , A. ; Aia , M. ; Min , J. ; Sichitiu , M.L ., 16 ( 3 ) , pp . 511-524 ., , 2008

3. Adaptive channel hopping for wireless sensor networks, Du, P., Roussos, G., 10-12 pp. 19-23, , 2011

4. Fusion in an Industrial Wireless Sensor Network for MCM, Kreibich , O. ; Neuzil , J. ; Smid , R. Quality-Based Multiple-Sensor, 61 ( 9 ) , pp . 4903- 4911 ., , 2014

5. Aggregation convergecast scheduling in wireless sensor networks, Nikolaidis, I., Malhotra, B., Nascimento, M. A., 17(2), pp. 319-335, , 2011

6. Aggregation convergecast scheduling in wireless sensor networks, Malhotra , B. ; Nikolaidis , I. ; Nascimento , M.A, 17 ( 2 ) , pp . 319-335 ., , 2011

7. Spatial TDMA: A Collision-Free Multihop Channel Access Protocol, Kleinrock, L., Nelson, R., 33(9), pp. 934-944, , 1985

8. TFMAC : Multi-channel MAC protocol for Wireless Sensor Networks ., Jovanovic , M.D . ; Djordjevic , G.L ., pp . 23-26 ., , 2007

9. MC-LMAC: A multi-channel MAC protocol for wireless sensor networks, Incel, Ö. D., Jansen, P., Hoesel, L. V., Havinga, P., 9(1), pp. 73-94, , 2011

10. Collision-Free Medium Access Control for Wireless Sensor Networks ., Rajendran , V. ; Obraczka , K. ; Garcia-Luna-Aceves , J.J. Energy-Efficient, 12 , pp . 63 ? 78, , 2006