第 112 回 碩士學位論文
指導敎授
趙
成
來
무선 인체통신 네트워크를 위한
복합 우선순위 MAC 기법
Hybrid Priority Medium Access Control
for Wireless Body Area Networks
中央大學校 大學院
컴퓨터공학 컴퓨터네트위크 專攻
董 冬
2010 年 2 月
무선 인체통신 네트워크를 위한
복합 우선순위 MAC 기법
Hybrid Priority Medium Access Control
for Wireless Body Area Networks
이 論文을 碩士學位論文으로 提出함
2010 年 2 月
中央大學校 大學院
컴퓨터공학 컴퓨터네트위크 專攻
董 冬
董 冬의 碩士學位論文으로 認定함
審査委員長 ㊞
審査委員 ㊞
審査委員 ㊞
中央大學校 大學院
2010 年 2 月
ACKNOWLEDGEMENT
The work contained in this thesis would not have been possible without
the support, encouragement and assistance of many people.
I would firstly like to thank my principal supervisor, Prof. Sungrae Cho,
for his commitment and support throughout the course of my research. His
professional advice, assistance and guidance have been greatly
appreciated. Prof. Cho has greatly contributed to my research and to the
preparation of this thesis.
Now, I would like to acknowledge and thank my parents and family. Their
words of encouragements support and prayers have helped, especially
through the number of more difficult and stressful moments of my
research. I thank them for their understanding, love, patience and support.
Without them my journey would have been formidable. A special
acknowledgement goes to my mother, father and my girl friend Fan Wen
Juan whose have inspired and motivated my work through endless words
of encouragement and love.
I also thank the many past and present seniors and juniors in my
department, Mr.
and Mr. , who gave me so many helps when I first
came in Seoul and also take me as a younger brother.
DONG DONG
2009
1
TABLE OF CONTENTS
ABSTRACT ................................................................................................................. 7
1. INTRODUCTION .................................................................................................... 8
1.1. Motivation and Overview ........................................................................... 8
1.2. IEEE802.15.6 Standardized Trends .......................................................... 11
1.3. Thesis Organization .................................................................................. 14
2. RELATED WORKS ................................................................................................. 15
2.1. MAC Problem in Emergency Packet Transmission ............................... 15
2.1.1 Hidden node problem .............................................................................. 16
2.1.2 Retransmission problem and transmission confirms problem ...................... 17
2.2. Priority Medium Access Control Method Comparative Analysis ............ 18
2.2.1 Priority-pulses scheme for priority-ensured MAC........................................ 18
2.2.2 Preemption-EDCA priority MAC scheme ................................................... 21
2.3. Comparison of Similar Priority MAC Scheme ........................................ 23
2.4. Proformance metrics for hybrid priority MAC in WBAN ....................... 24
3. BASIC IDEA .......................................................................................................... 25
3.1. The Fixed Time Slot Beacon Priority Medium Access for WBAN ......... 25
3.2. How to Make the Priority Transmission? ................................................. 26
3.3. Priority Fixed Time Slot Beacon MAC Protocol ..................................... 28
2
3.3.1 Network topology ..................................................................................... 28
3.3.2 Architecture .............................................................................................. 36
3.3.3 Data transfer to a coordinator ..................................................................... 29
3.3.4 Data transfer from a coordinator ................................................................. 30
3.3.5 Error recovery ........................................................................................... 31
3.3.6 Urgent transmission................................................................................... 33
3.3.7 Protocol flow chart .................................................................................... 34
4. PERFORMANCE EVALUATION AND SIMULATION RESULTS..................................... 35
4.1. Performance Evaluation ........................................................................... 43
4.1.1 Over view of NS-2 .................................................................................... 43
4.1.2 Simulation environment and parameter ....................................................... 44
4.2. Simulation Results .................................................................................... 47
5. CONCLUSION AND FUTURE WORKS ..................................................................... 54
REFERENCES ............................................................................................................ 55
3
LIST OF TABLES
Table 1. Characteristics of WBAN ......................................................................... 12
Table 2. Comparison of similar priority MAC protocols ....................................... 23
Table 3. Hardware environment ............................................................................. 45
Table 4. MAC Layer Parameters ............................................................................ 45
Table 5. Priority Data Parameters ........................................................................... 46
4
LIST OF FIGURES
Fig. 1
Wireless Body Area Networks structure ……………………...8
Fig. 2
Some relize application in WBAN ……………………………9
Fig. 3
Candidate frequency bands for WBAN ………………………13
Fig. 4
Timeline of standardization activities of IEEE802.15.6.……..13
Fig. 5
A hidden node scenario ………………………………………..16
Fig. 6
Emergency packets and their accompanying priopulses………19
Fig. 7
Structure of a priopulse ……...………………………………...20
Fig. 8
CP-EDCA Timing Design ………...…………………………..22
Fig. 9
CP-EDCA ……………………………………………………25
Fig. 10 Priority transmission ………………………………………...26
Fig. 11 Beacon structure ……………….………….…………………..27
Fig. 12 Beacon size…………………..………….……………………..28
Fig. 13 BAN star topology ……….…......…………....……………...29
Fig. 14 Device architecture ..........……………………………………32
Fig. 15 Data transmission from device to coordinator..………….…..30
Fig. 16 Data transmission from coordinator to device …………….31
Fig. 17 Same urgent data retransmission can lead to collision ……32
Fig. 18 Other urgent data must wait random backoff time……...……..32
Fig. 19 Urgent data transmission …............................................39
5
Fig. 20 Device transmission flow chart ……………………….……34
Fig.
21
………………………………..………….40
Fig. 22 Emergency data transmission …..……………………………40
Fig. 23 (a) Poll procedure for coordinator (b) Polling transmission flow
chart ………………………………..……………………...42
Fig. 24 Priority data dropping probability by number of data nodes ….47
Fig. 25 Average throughput by packets size …………………...…..48
Fig. 26 Priority data dropping probability by Packets by packet size …49
Fig. 27 Average throughput by number of data nodes …………...……50
Fig. 28 Average Delay by packets size …………………..……………51
Fig. 29
Collision probability by number of senders …………………..52
6
Hybrid Priority
Medium Access Control for
Wireless Body Area Networks
Computer Science and Engineering
Graduate School, Chung-Ang University
Seoul, S.Korea
Supervised by Professor Sungrae Cho
Abstract
The last few 10 years a wireless body area network (WBAN) can provide
a wide range of applications (e.p. non-medical service / Patient's urgent
state monitoring system..). Because of the restriction by some hardware’s
conditions. So we need to design the high-efficient MAC layer to improve
the transmission efficiency of wireless body area networks. On the other
hand, when the emergency situation appearing, a priority MAC
mechanism can send the urgent information to terminal service at first and
also can make sure that urgent information arrived. This thesis presents the
Hybrid Priority MAC Mechanism for wireless body area networks using
on-body device. Thus, when patients have an emergency situation allows
adequate assistance time and medical treatment time for patients. The
simulation based on NS-2, in the performance evaluation, our Hybrid
Priority MAC has the good performance and usability. Therefore, my
research in this area has high practical significance.
7
1.
Introduction
1.1. Motivation and Overview
The WBAN (Wireless Body Area Network) technology is still in its
primitive stage and is being widely researched. This technology once
accepted and adopted is expected to be a breakthrough invention in
healthcare leading to concepts like telemedicine and mHealth becoming
real [1]. Follows as the figure 1 we can see some application of WBAN.
Fig 1. Wireless Body Area Networks structure.
In the WBAN, each kind of application aspect, regarding patient's
guardianship, disease's prevention and the treatment is most essential. The
modern science's and technology's development causes the WBAN’s
application becomes diversifies. The Smart wireless electrocardiography
(ECG) patch, Body area network for wireless sleep staging, Body area
network for monitoring autonomic nervous system responses and
8
Autonomous wireless EEG monitor has already obtained the practical
application.[2]
Fig 2. Some realize application in WBAN.
WBAN is provides short range, low power and highly reliable wireless
communication for use in close proximity to or inside body. Typical
communication range is around 2 meters and which can be extended to 5
meters optionally. Various applications can be supported by WBAN.
IEEE 802.15.6 is the task group for BAN. The performance of a Wireless
Body Area Network (WBAN) to integration is required by miniaturized,
real-time transmitted, intelligent, in-body and on-body sensor nodes to
monitor body condition and the emergence of the urgent information.
9
In-body sensor can ocular get much information, because of the
limitations of cost, it is unfavorable to popularize [3].
Therefore, some emergency information’s transmission has become the
very essential question in WBAN. Because the In-body device has the cost,
difficulty level, signal weak and so on insufficient characteristic. So the
on-body monitoring device has many advantages (Easy to install,
Real-time monitoring, low cost, etc..) to popularize.[4] Also In-body
sensor node and on-body sensor node is used different channel to
communicate with terminal server.
In this paper, we proposed a new priority mac mechanism when used
on-body as it only using one channel. How can priority send the urgent
information first is an important issue. We synchronization monitoring a
wide range of information using one On-body sensor node , When the
patients has the emergency situation (Normally the sensor node must be
required within 1 second corresponding priority and will priority send an
emergency message, so patients can get adequate treatment time) using the
priority MAC can send the emergency signal immediately and the
emergency signal will be confirzd`1med again to received or not, If there
is no emergency signal was received, sensor node will send an emergency
signal again until they were confirmed. Thus, when patients have an
emergency situation allows adequate assistance time and medical
treatment time for patients. Therefore, research in this area has high
practical significance.
10
1.2. IEEE802.15.6 Standardized Trends
The standardization task group for wireless body area networks,
IEEE802.15.6, has started its activity since December 2007. Although
main focus of IEEE802.15.6 is laid on medical healthcare applications,
some members are looking at nonmedical applications such as consumer
electronics (CE) including entertainments.
On one hand, there is a possibility that both medical and non-medical
applications are able to be supported by a same PHY. On the other hand,
requirements on MAC, security, and QoS for medical and non-medical
applications would be different [5]. After intensive discussion on project
authorized request (PAR) and five criteria (5C) in several consecutive
meetings, the former study group summarized its uniqueness as shown in
Table 1.
Another hand, the selection of frequency bands for wireless BAN is
another important issue. There are existing frequency regulations for
medical applications. One of the regulations is wireless medical telemetry
(WMT).
The
definition
of
the
WMT
regulated
by
Federal
Communications Commission (FCC) is as follows.“ The measured
physiological parameters and other patient related information are
transmitted via RF communication between a patient-worn transmitter and
a remote monitoring unit.”
11
Table 1 Characteristics of WBAN
The Other 802
WBAN
Standards
MAC Layer
Single Scalable MAC
15.3 and 15.4 MACs
protocols
with Reliable Delivery
Power
Extremely Low to
Low
Consumption
protect Human
Power
Conventional
Conventional or with Body Energy
Source
Power Source
Scavenge Operation
Frequency
Medical Authorities approved
ISM or UWB
band
Communication Bands
Communication
Air, Vicinity of Human Body,
Air
Medium
and inside Human Body
FCC defines three frequency bands for WMT as shown in figure 3. They
are (1) 608-614 MHz, (2) 1395–1400 MHz, and (3) 1427-1432 MHz.
However, ITU recommends different usage on these frequency bands.
Therefore, other countries have different regulations from FCC. In Fig. 3,
WMT bands in Japan are also shown. [6] A common problem for WMT
and MIC is that bandwidth of a single channel is usually narrow in current
regulations. That limits high data rate applications.[7]
Until now, IEEE802.15.6 standardization task group is also carrying on a
series of standardized work (e.g. PHY, MAC, QOS, Secure, and
Device.Etc). Fig.4 shows the time line for BAN group in the near future.
12
Fig 3. Candidate frequency bands for WBAN.
Fig 4. Timeline of standardization activities of IEEE802.15.6.
Therefore, my research content has the very strong practical significance,
Started from last year, I followed closely of the standardization work and
Starts my research subject. I have referred to standardize the discussion
document so the document has the very big promoter action regarding my
aspect's research.
13
1.3. Thesis Organization
This thesis consists of five sections. In section 1 presents the wireless
body area networks standardized development and the application in
various fields especially guard the application in medical treatment. We
also introduced the composition of wireless body area networks.
Section 2 presents the related work. We compared to several kinds of
different priority medium access control mechanism. Under the different
environments to them, we research them how to realize the priority
transmission and their advantage and weakness. In wireless body area
network, because of possessing special properties. We propose the
requirements of some designs under the environments of WBAN and
Performance Metrics for Hybrid Priority MAC in WABN.
Section 3 provides the basic idea of this protocol, including 3 main parts.
First we proposed some cores of this thesis’s design. At the section 2
detailed expositions how did we realized the priority transmission
mechanism.
Finally,
we close
combination the
IEEE 802.15.6
standardized process and put forward the concrete MAC design plan. We
consider the completion of priority MAC from many views about WBAN
network topology, Architecture, Error recovery and so on. So prove the
practicability of our design form these respects.
The performance evaluation and simulation results are discussed in
Section 4. We also introduce our simulation scenario and parameters in
detail.
Finally, the conclusion and future works of this thesis are given in Section
5.
14
2. Related Works
In this section, we first explained some medium Access problem that often
appearing in WBAN. These problems are all very important respect to
design a priority MAC. And then we compared to several kinds of
different priority medium access control mechanism about Priority-pulses
MAC scheme, Preemption-EDCA priority medium access control scheme,
IEEE 802.11e, Traffic based MAC scheme and Narrow-band MAC
scheme. Under the different environments to them, we research them how
to realize the priority transmission and their advantage and weakness.
Finally we proposed Performance Metrics for Hybrid Priority MAC in
WABN.
2.1. Medium Access Control Problem in Emergency
Packet Transmission
2.1.1. Hidden node problem
Just like the Wireless body area networks and other wireless networks has
the same characteristic. The hidden node problem in WBAN is still one
very important, and needs to solve question. Because if we must make the
urgent signal fast transmission, hidden node problem is the question which
must solve.
The notorious hidden node problem deals with a configuration of three
nodes, like A, B, and C in Figure 5, whereby B is within the transmission
range of A and C, while C is outside the range of A. In a situation like this,
C will not be able to detect the ongoing transmission of A to B by carrier
15
sensing and, consequently, it can inadvertently interfere with B’s reception
of A’s packet.
The transmission range of a node A is defined as the area inside which
other nodes are able to correctly receive A’s packets. On the other hand,
the carrier sense range of A is the area encompassing those nodes whose
transmission A can perceive (carrier sense) while not necessarily being
able to receive the transmitted packets.
Fig 5. A hidden node scenario.
Generally, it is unreasonable to assume that the two areas are always the
same, e.g., the carrier sense range can be twice the transmission range [8].
Suppose that every node in Figure 5 has the same transmission range
(represented by a solid circle). Node C is out of the transmission range of
node A and thus would appear as a hidden node to A. However, if the
carrier sense range of C is larger than the transmission range of A (see the
dashed circle), C is hidden no more because it can sense the transmission
16
of A and thus avoid interfering with it. This mechanism for eliminating the
hidden node problem has been described in [8].
Carrier detection is usually controlled by thresholds applied to the level of
(actually or apparently) perceived signal. Low thresholds tend to be
sensitive to many factors involving more than the distance between the
nodes, e.g., the natural noise level in the neighborhood. While formally,
increasing the carrier sense range is possible, low thresholds may trigger
many false indications, which will result in unnecessary back-offs and
reduced throughput, possibly below one that could be achieved by simply
ignoring the hidden node problem altogether.
To alleviate the hidden node problem, Karn [9] proposed a two-way
handshake involving short packets whose exchange should precede the
actual transmission. The sender starts by transmitting a Request-To-Send
(RTS) packet. After receiving RTS, the intended recipient sends a
Clear-To-Send (CTS) packet to the sender. Both packets specify the length
of time needed to transmit the actual data packet. Any third party node
receiving any of the two packets will know for how long it should refrain
from transmission as to avoid interfering with the exchange in progress.
This protocol has been standardized into the popular IEEE 802.11 family
of access schemes. The complete exchange involves four packets:
RTS/CTS/DATA/ACK, with the first pair taking care of the hidden nodes,
and the final ACK providing for reliable delivery (triggering
retransmissions on failures).
17
2.1.2. Retransmission Problem and Transmission Confirms
Problem
Retransmission has been adopted as one of the important scheme for
improving transmission reliability in wireless body area networks. When
the emergency situation appearing, in emergency information transmission
processing, if appearing the delay or packet lost, we must be let the
information retransmit by the priority MAC protocol. Then determined
whether transmits successfully is a very essential aspect. So if we must
solve retransmission problem that to first consider whether can determine
transmits successfully or transmission confirms. Therefore these two
problems are close related.
2.2. Priority Medium Access Control Method
Comparative Analysis
2.2.1. Priority-pulses MAC scheme for priority-ensured
medium access
The basic approach of the proposed MAC scheme is to use “pulses” in a
single control channel to achieve multiple goals. This scheme has
proposed by “Jun Peng” .The “pulses” in this scheme are basically
single-tone waves with pauses of random lengths. These New MAC
protocols focus on providing statistical priority for unicast flows instead of
strict priority for individual packets. With its novel pulse-based control
mechanism, This MAC scheme realizes strict packet-level priority
scheduling for emergency packets in a fully distributed way [10]. I will
carry on the introduction to this mechanism.
18
Follows as figure 6 to introduce how to realize the priority transmission.
Node A is the emergency message source, node B is a neighbor of node A,
and node C is a hidden terminal to node A. The signals in the control and
data channels are shown.
When a node is using the data channel for disseminating emergency
packets, it continuously transmits priopulses in the control channel, the
node stops its priopulses only after its transmission of emergency packets
is finished or other emergency message sources interrupt it. Priopulses not
only ensure that an emergency packet receives the actual priority that it
deserves for its level of emergency but also suppress the hidden terminals
Fig 6. Emergency packets and their accompanying priopulses.
when the emergency message is in transmission. Each priopulse consists
of an active part of fixed length and a pause part of random length, as
shown in Fig. 7. In the active part, single-frequency waves are transmitted
in the control channel, while there is no wave transmitted in the pause part.
The random pause part is further divided into two parts, i.e., a contention
window of fixed size and a residual pause of random length. Moreover,
the contention window is cut into equal-size subwindows.
19
Node A is the emergency message source, node B is a neighbor of node A,
and node C is a hidden terminal to node A. When the emergency message
must be transmit by node A, but node B is using the channel, so node A
must be wait the one backoff time until the one priopulse is finished the
active part. So node A can use the channel to transmit emergency message.
Takes in the channel period, node C always idle the channel. Before but
when node A prepares uses the channel, he essential waited for
sub-windows backoff time to determine whether have a higher priority
node. The Residual Random pause to solve the problem that is two same
priority nodes which one is use the channel first.
This new MAC mechanism's can succeed solution migration node priority
problem, hidden node problem and can enhance the network transmission
speed, reduces the network delay because of the priority competition. But
this mechanism has the big network expenses and fit on the mobile nodes,
therefore this new mechanism does not adapt in wireless body area
networks. Because in WBAN basically is the relatively stationary position
nodes.
Fig 7. Structure of a priopulse.
20
2.2.2. Preemption-EDCA priority medium access control
scheme
Manikanden Balakrishnan [11] has proposed the priority MAC scheme
that used channel service preemptions during random medium access. In
the context of time-critical sensor applications, emergency traffic will
have the privileges to interrupt the services of other routine traffic in the
network to guarantee the lowest possible channel access latencies. This
scheme design within the QoS framework of the popular 802.11e EDCA
standard and demonstrate the importance of service preemptions for
emergency
reporting.
The
proposed
Channel
Preemptive-EDCA
guarantees the lowest delay bounds and immunity to the presence of other
lower priority traffic, for emergency frames, even under high network
loads. The cost, however, is increased delays for routine traffic, which is
acceptable during emergency sensor reporting.
Fig 8. CP-EDCA Timing Design.
Figure 8 depicts the timing relations, which are crucial for this preemption
design. The basic concept behind this formulation is similar to the IFS
design of 802.11 schemes, which uses the wait times to allow for
21
precedence in acquiring a free channel. For example, in 802.11e EDCA an
on-going frame bursting sequence (Data-Ack-Data) is separated by SIFS,
which implies that it takes the highest precedence irrespective of the frame
priority [11].
Fig 9. CP-EDCA
Follows as figure 9 detailed explanation CP-EDCA design mentality,
guaranteeing medium access for emergency priority through TXOP
preemptions. Emergency frame sequence is separated by the shortest wait
time, new emergency frames start channel acquisition or contention if they
sense the channel free for an EPAIFS period, which is smaller than
NPSIFS. The data-ack sequence is never interrupted and can be achieved
by setting the Network Allocation Vector to the end of ack transmission.
22
Although this service preemptions priority medium access scheme has
solution the priority transmission problem. But there has the obvious
shortcoming. First in wireless body area networks for longer time work the
preemptions scheme will takes much memory and energy. Such words, the
nodes operating time cannot achieve the long times working request.
Second, this scheme can not to solve the hidden node problem in wireless
body area networks. Preemptions by hidden nodes will be getting bad
result for emergency information transmission owing to collisions. Third,
because of channel errors on QoS guarantees, since frame errors and
re-transmissions will be have a delay performance by this preemptions
scheme.
2.3. Comparison of Similar Priority MAC Schemes
Table 2
Comparison of Similar Priority MAC Protocols
Priority
Service
Traffic-based
Narrow-band
Pulses MAC
Preemption
IEEE
MAC(IEEE
MAC
(IEEE
MAC(MILCOM
802.11e
802.15.6
(IEEE 802.15.6
VTS 07)
08)
Presentation)
Presentation)
high
high
normal
normal
normal
high
low
normal
high
high
Acknowledgment
yes
yes
yes
yes
yes
Delay
high
normal
normal
normal
low
MAC
Protocols
Operations
Energy
Consumption
Memory
Consumption
23
Though to analysis, the above several kinds of priority MAC mechanism
our hybrid priority MAC is quite simple structure and the energy
consumption is low. Also can solution the urgent delay problem,
acknowledgment problem and memory consumption problem. Although
they other priority MAC will be possibly more perfect in the performance
but our MAC even more being suitable in wireless body area networks.
Below I will talk bout the performance metrics in our design.
2.4. Performance Metrics for Hybrid Priority MAC in
WABN
We defined performance metrics for hybrid priority MAC in WBAN as
follows:
The simplicity is an important part in my MAC design, because of the
wireless body area networks emphasis is the simple usability and can
provide the long time stable use. The simple reliable design can let the
wireless nodes achieve the maximum efficiency. Low latency for urgent
data also is aspect which must consider for WBAN priority MAC design.
Recharging sensor node batteries can be inconvenient or impossible that is
why wireless body area networks should function for as long as possible.
Therefore protocols must be designed to be low energy efficient. [13] My
protocol’s most important part is the guarantee urgent data priority
transmission and confirms the urgent data transmission. Because of we
must guarantee the patient can obtain the rescue in the shortest time, if the
protocol lost or delay emergency information that means we can not use
this protocol in wireless Body area networks. High reliability also will
consider in our priority MAC protocol.
24
3. Basic Idea
3.1. The Fixed Time Beacon cycle Priority Medium
Access for Wireless Body Area Network.
My thesis proposed the priority medium access is based on the fixed time
slot beacon cycle and the different priority level uses the different back-off
time slot. In our beacon based access, Device sends the data to coordinator
uses CSMA/CA. But the coordinator sends the data to device uses TDMA
to provide the low energy consumption.
Takes this kind of hybrid priority way, there are two reasons that we using
the fixed time slot beacon cycle. First of all, because the emergency
situation will not appear at any time. If the time slot of cycle is too long,
then when taking place in emergency, it can't fast transmit the emergency
information. This has caused priority transmission will be failed. Secondly,
the energy consumption is a respect having to be considered in wireless
body area networks. In order to make the wireless nodes work long times,
we using the rational time slot beacon cycle to prevent the surplus energy
consumption. These two above-mentioned points made me using the fixed
time slot beacon cycle. Therefore I define the first question, now I will
prove how to make the priority transmission.
3.2. How to Make the Priority Transmission?
In this thesis the point is how to transmit emergency information fast that
is to say how to realize the priority transmission. p
25
Fig 10. Priority transmission.
Follows as the figure10 during the each time of the beacon cycle, we use
interval of different backoff time slot to be sent the high priority data first.
While the high priority data are sent, the general data must avoid making a
backoff. This time slot is NIFS. It’s over to only reach the high priority
data has transmission after receiving to ACK and confirm it and then the
low priority data begin to send. The figure10 has recommended solving
the priority transmission from microcosmic. Through the figure 11 I will
introduce the polling data transmission course from macroscopic.
Before each beacon cycle time plans to begin, the coordinator will be
using the fixed cycle time slot to control the nodes which is urgent node
and which is low priority node. The data produced by these different
priority nodes will use different intervals within one cycle plans. After one
beacon cycle plan finished, the coordinator will wait the fixed time slot
and begin next beacon. Then follows as the figure 11 shows the each
beacon structure. Each beacon includes the next beacon time Transmission
time (Coordinator->Device) and Receive time (Empty Slot for each device
26
to access the slot in CSMA/CA). Transmission Time and Receive Time
are exploited to save energy. Let the device Sleep and Transmission.
Fig.11. Beacon structure
How bout beacon cycle time slot?
In our designs, the beacon cycle time slots are appears by the fixed form
on a period of time. When two beacon time intervals are excessively long,
there cannot guarantee that the urgent data transmits fast. When two
beacon time slot intervals are excessively short, there will present the data
collision. How can coordinate the size of beacon time slot is a key
question. To investigate the maximum throughput (S) and the mean
beacon period, we assume there is only one data frame in each queue all
the time (saturation condition). The mean beacon period is defined as the
mean time between two consecutive new beacon turns of a station. Where
n is the number of stations in the list. The wireless link channels between
the coordinator and devices are assumed to be Omni and independent of
each other. Follows as the figure 13 shows the each beacon size and if
traffic is too much, short Beacon cycle is not enough to cover traffic. So,
Beacon cycle is adaptively adjusted depending on current traffic.
27
Fig.12. Beacon size
3.3. Fixed Beacon cycle Priority MAC Protocol.
3.3.1. Network topology
To meet the application requirements, an IEEE 802.15 TG6 - Body Area
Network (BAN) may operate in star topologies or extended star topologies.
My priority MAC design is based on the star topology; however the
proposed solution has a scope to expand it to extended star topology in
future [1]. In a star topology, as shown in Figure 13, the communication
session is established between an end device and a BAN Coordinator. For
on-body communication, both coordinator and device can initiate or
terminate the communication, additionally coordinator can route data from
one device to another device. For implant communication, device can not
initiate communication except in occurrence of an emergency event at
device. In BAN, primarily, a device generates traffic related to one
application. Coordinator may or may not generate traffic related to an
application.
28
Fig 13. Ban star topology.
3.3.2. Architecture.
Figure 16 illustrates the Architecture for WBAN device. An IEEE 802.15
TG6 device may contain PHY1 or PHY2 or both PHY1 and PHY2, which
contains transceiver for signal transmission and reception. The PHY1
transceiver operates in a frequency band suitable for impl ant
communication and PHY2 transceiver operates in a frequency band
suitable for on-body communication. An IEEE 802.15 TG6 device also
contains a MAC and LLC layer to access a channel of a selected
frequency band for all kind of data transfer. [14] In our design we just
only consider the situation of on-body communication and use the PHY-2.
29
Fig 14. Device architecture.
3.3.3. Data transfer to a coordinator.
In our beacon based access, Device sends the data to coordinator uses
CSMA/CA. In each beacon, provides the contention way to guarantee the
urgent data to transmission first.
Fig 15. Data transmission from device to coordinator.
3.3.4. Data transfer from a coordinator.
In our beacon based access, Coordinator sends the data to device uses
TDMA to provide the low energy consumption.
30
Fig 16. Data transmission from coordinator to device.
3.3.5. Error recovery.
In order to provide reliable packet transmission the standard supports
beacon based error recovery (coordinator driven) applicable only to
upstream traffic with beacon based channel access. the beacon based error
recovery is provided to for highly power constraint devices and does not
allow retransmission of packet without any response received from the
coordinator otherwise it can lead to the collision of same priority packet
with one beacon cycle. The situation is shown in Figure 19.
The beacon based error recovery is making us design a respect considered.
In our mechanism if node has same urgent data that coordinator must
receive. Then the node must wait for random backoff time slot that can be
transmission. The situation is shown in Figure 20.
31
Fig 17. Same urgent data retransmission can lead to collision.
Fig 18. Other urgent data transmission must wait random backoff time.
32
3.3.6. Urgent transmission.
Emergency packet transmission is one of the most crucial requirements in
any telemedicine systems. However, urgent messages have highly erratic
nature; it has to be sent as soon as possible. The urgent data should not be
delayed or denied due to dynamic availability of network resources. My
thesis’s purport is to solve the urgent data priority transmit problem. We
mention in section two of this chapter detailed how to solve this problem.
So we just will show the main part of the fixed beacon cycle time slot
priority MAC protocol.
Fig 19. Urgent data transmission.
Follows as figure 22 when the urgent data appearing, during the each time
of the beacon, we use interval of different backoff time slot to be sent the
33
high priority data first. While the high priority data are sent, the general
data must avoid making a backoff. This time slot is NIFS. It’s over to only
reach the high priority data has transmission after receiving to ACK and
confirm it and then the low priority data begin to send. Then the high
priority node will be idle until the g node have already completed the RTS,
CTS, DATA and ACK. At this time the next beacon time slot will be start.
3.3.7. Protocol flow chart.
Now I will show slow chat of our fixed time slot priority poll mechanism.
Follows as the figure 23 will prove how MAC worked from start to
finishes.
Fig 20. Device transmission flow chart.
34
4. Performance Evaluation and Simulation
Results
4.1. Performance Evaluation
In this section, for evaluating the benefits of our Hybrid Priority MAC, we
compare the simulation results with traditional IEEE-802.15.4 and IEEE802.11e. We first implement the proposed IEEE 802.11e TKN MAC in
NS2-2.28 version [18], And then, we have modify the IEEE-802.15.4
MAC protocol (this simulator made by SAMSUNG and CUNY [19]) to
compare these protocols in different aspects. We in view of the priority
transmission’s problem altogether have made five parts of Performance
evaluation. These includes: Urgent/Normal data access delay, Average
throughput, Average delay, Collision probability and Energy consumption.
The simulations are done by assuming a star topology with the master
node as the coordinator. Because the physical-layer of the BAN is still
open to be designed, in our simulations the physical layer parameters are
defined according to the IEEE 802.15.4 standard.
4.1.1. Overview of NS-2
NS-2 is a Net work Simulator, built with C++ and TCL, some standard
algorithms are already implemented in this simulator. As every simulator,
the main purpose is to simulate different networks, to test different
protocols, and to find the limitations of each. It has been developed in the
35
California University, by LBL, Xerox PARC, UCB, and USC/ISI through
the VINT project supported by DARPA.
The NS-2 makes use of flat earth model in which it assumes that the
environment is flat without any elevations or depressions. However the
real world does have geographical features like valleys and mountains.
NS2 fails to capture this model in it. Many researchers have proposed the
additions of new models to NS2. Shadowing Model in NS2 attempts to
capture the shadow effect of signals in real life, but does that inaccurately.
NS2's shadowing model does not consider correlations: a real shadowing
effect has strong correlations between two locations that are close to each
other. Shadow fading should be modeled as a two dimensional log-normal
random process with exponentially decaying spatial correlations. [15]
4.1.2. Simulation Environment and Parameter
The simulation experiment environment is divided into hardware
environment and software environment. Hardware environment is follows
as the table 3.
The IEEE 802.11e TKN model and CUNY 802.15.4 model were used in
our simulation. We have used two kinds of modes to differentiate the
accuracy of the simulation. We also having used the transmit nodes
quantity and the size of data packets to comprehensive appraisal. In the
nodes arrangement respect, we have used the mode by fixed position,
because in common wireless body area networks the nodes distance is
very near and the position is fixed.
We have used the wireless channel. The Radio propagation model is
Two-Ray ground model. Also the Antenna type is Omni-Antenna. We
36
have set up a wired node in centre. Because of the WBAN can not be more
nodes [16], while using the nodes quantity model we just set the value is
#3---#10. While using the size of data packets model we set the value is
10(Bytes), 20(Bytes), 40(Bytes), 60(Bytes), 80(Bytes), 100(Bytes) …
800(Bytes).
Table 3 Hardware Environment
CPU
Inter Core 2 Duo 2.8 GHz
RAM
4G
OS
Cygwin (Linux)
NS-2 Version
2.28
802.11e
TKN Module
Module
802.15.4
CUNY LR-802.15.4
Module
In MAC layer the parameters is follows as table 4.
Table 4 MAC Layer Parameters
Parameters
Traffic Type
Packet Size
Data Interval
Topology
Routing Type
Antenna
BO value
Value
CBR
15 Bytes
0.001ms
Star Topology
AODV
Omni
2
37
SO value
Ifs
Sifs
Uifs
Nifs
Cwmin
2
16us
8us
6us
72us
16
Cwmax
1024
Urgent CW
7-15
Normal CW
31-1024
The urgent data and normal data parameters follows as in the priority
model that we set up:
Table 5 Priority Data Parameters
Urgent data
Normal data
Prio 0
Prio 1
Prio 0 AIFS 2
Prio 1 AIFS 2
Prio 0 CW_MIN 7
Prio 1 CW_MIN 31
Prio 0 CW_MAX 15
Prio 1 CW_MAX 1024
According to different CW size, the data with high priority will get the
least back-off time.
38
4.2. Simulation Results
We have used the range of 10m*10m. Follows as the figure 24 shows the
nodes and coordinator produced under NS-2.
Fig 24. Coordinator and nodes in WBAN.
For typical number of nodes used in WBAN [20]. In order to provide the
urgent data priority transmission. We had performance evaluation for
urgent data average access delay as show in figure7. We trough the
number of data nodes to consider the urgent data average access delay
between HP-MAC, IEEE-802.15.4, and IEEE-802.11e. The number of
nodes from 2 to 18. Our HP-MAC can provide there are very low delay
rates for urgent data. Through the increasing number of nodes, HP-MAC
has smaller delay increase rate. It means HP-MAC can priority transmit
the urgent data.
39
Urgent data average access delay (ms)
6
HP-MAC
IEEE 802.15.4
HP-MAC (max value)
IEEE 802.15.4 (max value)
5
4
3
2
1
0
2
4
6
8
10
12
Number of data nodes
14
16
18
8
Urgent data average access delay (ms)
7
6
5
4
3
2
HP-MAC
IEEE 802.11e
HP-MAC (max value)
IEEE 802.11e (max value)
1
0
2
4
6
8
10
12
Number of data nodes
14
16
18
Fig 25. Urgent data access delay by number of data nodes.
Because the normal data must wait long back-off time. Even though
HP-MAC has higher delay when normal data transmission. HP-MAC can
provide urgent data priority transmission. Until to analysis max delay for
HP-MAC, HP-MAC has low difference for max delay that means
40
HP-MAC has low delay jitter. This situation appears in figure 8. In the
standardized regulation of the IEEE-802.15.6, the delay of the normal data
(e.p. ECG, Hearing. etc...) can't exceed 250ms. Through the increasing
number of nodes, HP-MAC has lower increase rate too. This situation to
prove our design is suitable for WBAN.
60
HP-MAC
IEEE 802.15.4
HP-MAC (max value)
IEEE 802.15.4 (max value)
Normal data average access delay (ms)
50
40
30
20
10
0
2
4
6
8
10
12
Number of data nodes
14
16
18
2
Normal data average access delay (ms)
10
1
10
HP-MAC
IEEE 802.11e
HP-MAC (max value)
IEEE 802.11e (max value)
0
10
2
4
6
8
10
12
Number of data nodes
14
16
18
Fig 26. Normal data access delay by number of data nodes
41
On one hand, follow as the figure 9 we trough the number of data nodes to
consider the average delay between HP-MAC, IEEE-802.15.4 and
IEEE-802.11e In order to appraise the performance of HP-MAC, We need
to consider the whole performance. The number of nodes from 2 to 18.
According to the simulation result, our HP-MAC and IEEE-802.15.4 are
overall performance difference is small, because of this respect, though
HP-MAC are big on the average delay, but HP-MAC can totally guarantee
to urgent data transmit very fast.
30
HP-MAC
IEEE 802.15.4
Average delay (ms)
25
20
15
10
5
0
2
4
6
8
10
12
Number of data nodes
14
16
18
14
16
18
50
HP-MAC
IEEE 802.11e
Aaverage delay (ms)
40
30
20
10
0
2
4
6
8
10
12
Number of data nodes
42
Fig 27. Average delay by number of data nodes
On the other hand, follow as the figure 10 we trough the number of data
nodes to consider the average throughput between HP-MAC and
IEEE-802.15.4. The number of nodes from 2 to 18. When data nodes
number is 4, our HP-MAC average throughput is 115kbps and the
IEEE-802.15.4 MAC average throughput is 129 kbps. Until the nodes
number rises to 18, IEEE-802.15.4 is still has higher throughput than HPMAC. Because in preceding simulation, though the influence of average
delay, HP-MAC does not have preponderate, because in order to ensure
the urgent data fast transmission, we can only sacrifice the whole
performance for HP-MAC.
140
HP-MAC
IEEE 802.15.4
130
Throughput (Kbps)
120
110
100
90
80
70
60
2
4
6
8
10
12
Number of nodes
14
16
18
Fig 28. Throughput by number of nodes
43
This part of simulation is about the urgent data and normal data’s
collision probability analysis. The number of sender nodes from 5 to 30
and takes 5 nodes to increase. Follows as the figure 11 when the sender
nodes number is 5, the collision probability of urgent frame using
HP-MAC is %0.14, the collision probability of normal frame using
HP-MAC is %0.51, and the collision probability of urgent frame using
IEEE-802.15.4 is %0.21, and collision probability of urgent frame using
IEEE-802.11e is %0.49. This indicated that our HP-MAC has the small
collision probability to lead in priority transmission aspect. And along
with number of nodes increasing, our HP-MAC collision probability
scopeis smaller than 802.15.4 and 802.11e. This proves, HP-MAC has
very strong reliability for urgent data transmission. The advantage of our
scheme lies in ensuring the sending urgent quickly and safety.
1
Urgent Frame HP-MAC
Urgent Frame IEEE 802.15.4
Urgent Frame IEEE 802.11e
Normal Frame HP-MAC
Normal Frame IEEE 802.15.4
Normal Frame IEEE 802.11e
0.9
Collision Probability
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
5
10
15
20
The number of the senders
25
Fig 29. Collision by number of senders
44
30
Energy efficiency is a very important measurement in wireless sensor
networks [11]. For BAN, this is even critical, because the implanted
medical sensor nodes might not be rechargeable. Hence, the energy
efficiency enhancement becomes the key design target of the MAC
protocol for BAN. In simulation, for typical number of nodes used in
WBAN, with a high packet arrival rate, 10 packets per second, the number
of nodes from 6 to 18 and takes 2 nodes to increase. Our HP-MAC can
provide there are low energy consumption rates for data transmission
about 0.392Uj per kilo bits. But the IEEE-802.15.4 increase tate is higher
than HP-MAC. Through the increasing number of nodes, the
IEEE-802.11e has highest energy increase rate. Because all nodes of
IEEE-802.11e are real-time working, So 802.11e has very high energy
consumption, this point is not suitable for WBAN. Consider from energy
consumption, HP-MAC is very suitable for WBAN than IEEE-802.15.4
and IEEE-802.11e.
1.7
Average power consumption per kilo bits (μJ)
1.6
HP-MAC
IEEE 802.15.4
IEEE 802.11e
1.5
1.4
1.3
1.2
1.1
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
6
8
10
12
14
Number of nodes
16
18
Fig 30. Power consumption by number of nodes
45
Through to analysis of integral simulation experiment, when data quantity
and nodes number relatively increase, our HP-MAC have certain
shortcoming. Because our aim at the design a priority MAC for wireless
body area networks, so in being suitable in the WBAN our priority
mechanism obtained the good effect. Through the four parts, Priority data
dropping probability, Average throughput, Average delay and Collision
probability, the simulation experimental result proves our MAC
mechanism to be able the faster transmission urgent information, and has
certain usability in WBAN.
46
5. Conclusion and Future Works
In this thesis we present the Hybrid Priority MAC mechanism for wireless
body area networks using on-body device. Thus, when patients have an
emergency situation allows adequate assistance time and medical
treatment time for patients. The results from the simulation our Hybrid
Priority MAC mechanism can effectively solve such as sent the urgent
signal immediately and reduce the transmission delay. Through the hybrid
form can effective suppression collision in wireless body area network.
Through to analysis of integral simulation, when data quantity and nodes
number relatively increase, our HP-MAC have certain shortcoming. (E.g.
Delay Jitter, high-capacity data throughput weak …). But our priority
MAC mechanism to be able the faster transmission urgent information,
and has certain usability in WBAN.
In future work the priority MAC for WBAN is still a hot topic and
constantly have new device also use a different transmission mechanism.
In WBAN most applications due to medical care, so it is necessary to use
a MAC that has reliable to quickly send the urgent message. As a result of
this research continued to deepen, it will have more extensive application
of WBAN
47
6. References
[1] IEEE 802.15.6 TG6 Work Group.
[2] IMEC research center.
[3] IEEE 802.15 WPAN WG homepage, http://www.ieee.org/15.
[4] BAN project Authorization Request (PAR) draft,
IEEE P802.15-07-05750-06.
[5] BAN application matrix, IEEE 802.15-07-0735-00-0ban.
[6] Huan-Bang Li; Takizawa, K.; Kohno, R.; “Trends and standardization
of body area network (BAN) for medical healthcare” In Proc. EuWiT
2008. European Conference on 27-28 Oct. 2008.
[7] A. W. Astrin, H.-B. Li, and R. Kohno, “SG BAN project draft 5C,”
IEEE 802 15-06-0488/r7, July, 2007.
[8] K. Xu, M. Gerla, and S. Bae. “How effective is the IEEE 802.11
RTS/CTS handshake in ad hoc networks? ” In Proc.IEEE GlOBECOM,
volume 1, pages 17–21, November 2002.
48
[9] P. Karn. “MACA–a new channel access method for packet radio”. In
Proc. Computer Networking Conference on ARRL/CRRL Amateur
Radio, pages 134–140, September 1990.
[10] Jun Peng, Liang Cheng; “A Distributed MAC Scheme for Emergency
Message Dissemination in Vehicular Ad Hoc Networks” IEEE
TRANSACTIONS ON VEHICULAR TECHNOLOGY, VOL. 56, NO.
6, NOVEMBER 2007.
[11] Balakrishnan, M.; Benhaddou, D.; Xiaojing Yuan; Gurkan, D.;
“Service preemptions for guaranteed emergency medium access in
Wireless Sensor Networks ” In Proc. IEEE Military Communications
Conference, 2008.
[12] Qaimkhani, I.A.; Hossain, E.; “Efficient silence suppression and call
admission control through contention-free medium access for VoIP in
WiFi
networks”
Communications
Magazine,
IEEE
Volume
46, Issue 1, January 2008.
[13] Hoefel, R.P.F.; “Dynamic time slot scheduling schemes for uplink
polling MAC TDD DS-CDMA protocol with adaptive antennas”
IEEE Electronics Letters
Volume 38, Issue 19, 12 Sep 2002.
[14] IEEE Standards Department, “Wireless LAN medium access control
(MAC) and physical layer (PHY) specifications,” IEEE standard
802.11_1997, 1997.
49
[15] Network Simulation 2 (NS-2): http://www.isi.edu/nsnam/ns.
[16] Victor Shnayder, Bor-rong Chen, Konrad Lorincz, Thaddeus R. F.
Fulford-Jones, and Matt Welsh., “Sensor Networks for Medical
Care”, Harvard University Technical Report TR-08-05, April 2005.
[17] Taekon Kim, Hyungkeun Lee, Jinhwan Koh, Kyung-suk Lhee., “A
performance analysis of polling schemes for IEEE 802.11 MAC
over the Gilbert–Elliot channel” International Journal of Electronics
and Communications Volume 63, Issue 4, 3 April 2009.
50