Wireless Sensor Network, 2011, 3, 61-72
doi:10.4236/wsn.2011.32007 Published Online February 2011 (http://www.SciRP.org/journal/wsn)
A New MAC Protocol for Moving Target in Distributed
Wireless Sensor Networks
Jijun Zhao, Xiang Sun, Zhongcheng Wei, Zhihua Li
School of Information & Electronic Engineering, Hebei University of Engineering, Handan, China
Received November 29, 2010; revised December 16, 2010; accepted February 21, 2011
Abstract
A wireless sensor network (WSN) is composed of many nodes with limited power supply; most nodes are
stationary in the network which could probably involve a few mobile nodes. Various medium access control
(MAC) protocols specially aimed at a target locating application for WSNs have been proposed. However,
most of these protocols based on the problem of energy-limited does not consider the mobility of nodes.
Therefore, in order to solve such problem, this paper proposes a MAC protocol—Distribute Moving-MAC
(DM-MAC). Under the condition of keeping high network coverage, the protocol utilizes the redundancy of
nodes to strengthen the robustness and energy efficiency of network, and decreases the packet loss rate of the
mobile node’s communication links for realizing reliable communication of two nodes. Simulation indicates
that the new protocol has higher energy efficiency, lower packet loss rate and higher network coverage
which suit for wireless sensor network with mobile nodes well.
Keywords: Wireless Sensor Network (WSN), Distribute Moving-MAC (DM-MAC), LDAS, EAR, Mobility
1. Introduction
A wireless sensor network (WSN) has characteristics of
self-organization, multi-hop, large scale of monitoring
and energy constraints. Recently, many research institu-
tions have studied on this area and have a great break-
through on it. Because WSN has a strong ability of pro-
cessing ability within network, it has been applied to
various situations, such as battlefield monitoring, ma-
chine fault monitoring, biological monitoring, intelligent
furniture and so on. MAC protocol in WSN is mainly to
realize the reliable communication of point-to-point
(node to node), sleeping mechanism, synchronization
mechanism and so on. Therefore, MAC protocol has a
decisive function to whole network due to a great influ-
ence on energy efficiency, scalable to node density,
frame synchronization, fairness, bandwidth utilization,
flow control, and error control for data communication.
The most key feature in WSN is the limitation of node
energy, therefore, most MAC protocols have focused on
the energy saving in WSN without considering the mo-
bility of nodes. Nevertheless, there have been introduced
a certain amount of mobile node in many position sys-
tems and environmental monitoring systems. Thus, in-
creasing the reliability of the mobile nodes communica-
tion under the condition of ensuring network connection
and coverage becomes a mayor research trend of MAC
protocol study. In recent researches, Traffic-Adaptive
Medium Access protocol (TRAMA) [1] and WiseMAC
protocol [2] solve node sleeping mechanism well and
limit energy consumption of the node in a certain low
value, but bring about a large packet loss rate after sev-
eral mobile nodes being added into network. The
Lightweight Deployment-Aware Scheduling (LDAS)
algorithm proposed in [3] solves the problem of low
coverage (including sensor coverage and communication
coverage) which caused by nodes’ sleeping mechanism;
however, it doesn’t have a mechanism to realize the re-
liable communication, so it doesn’t accomplish dynamic
communication between the different nodes. The Eave-
sdrop-And- Register (EAR) algorithm proposed in [4]
which based on the Self-organizing Medium Access
Control for Sensor networks (SMACS) algorithm can
realize the reliable communication between mobile node
and fixed node by depending on increasing the control
head packets, but when adding the number of mobile
nodes, it results in the packets collision with a high
probability rate that leads to the node energy consump-
tion much faster and it also doesn’t guarantee high rate
of coverage in the monitoring area. The mobility-aware

J. J. ZHAO ET AL.
62
MAC protocol for sensor networks (MS-MAC) proposed
in [5] can adjust the scheduling mechanism dynamically
acco·rding to velocity of the node, but this protocol is
based on the Senor-MAC (SMAC) protocol in [6,7],
which doesn’t consider the problem of reliable commu-
nication of mobile node and the network coverage. Here
the network refers to this paper is the areas that are col-
lectively covered by the set of all sensors.
Based on the above researches, we find that the exist-
ing MAC protocols are almost proposed based on the
problems of energy consumption, network coverage and
network connection. In the recent researches, very few
MAC protocols could be used in a mobile network. Al-
though a few protocols could support a number of the
mobile nodes, they also exist the problem of high energy
consumption, limit number of mobile nodes and low
network coverage, etc [8-14]. DM-MAC proposed in the
paper is based on the network environment which exist
many stationary nodes and several mobile nodes. The
protocol decreases occurrence probability of blind hole
and solves unreliable point-to-point communications
which are caused by LDAS algorithm, and improves the
communication mode in EAR algorithm and intelligent
scheduling mechanism in MS-MAC protocol to realize
the communication between mobile nodes and stationary
nodes. Meanwhile, the protocol could ensure the network
having a good performance in network coverage and
network connection which also enhance robustness of the
network.
In the network, the function of stationary node is to
collect and monitor sensor data of surrounding environ-
ment, locate the geographical position of the mobile
nodes and send collecting data to sink node. As the target
node need to be located, mobile node is to broadcast
signals for the purpose of location by stationary node;
meanwhile, mobile nodes also collect environmental data
and send them to sink node by the stationary. Generally,
mobile node is always composed of stationary node
which fixed on the movable objects. We consider that the
limitation of energy and hardware is not a key problem
of the mobility, so some management tasks could be
given to mobile nodes for achievement.
The paper is organized as follows. In Section 2, we
analysis the existing protocols which adapt to the distri-
bute network for moving target; while Section 3 propos-
es a new protocol whose performance is superior to the
other existing protocols. Simulation results are presented
in Section 4 and Section 5 concludes our work.
2. Related Protocols Analysis
According to the aspects of sleeping mechanism, net-
work coverage and building communication link, etc, we
mainly analyze the following three protocols, with com-
paring their advantages and disadvantages for the future
research.
2.1. EAR Algorithm
The basic idea of EAR is to build reliable communica-
tion link though adding handshaking times and using
SNR value of receiving signals. In the EAR algorithm,
links are build up between mobile nodes and stationary
nodes considering four kinds of frames: Broadcast Invite
(BI), Mobile Invite (MI), Mobile Response (MR) and
Mobile Disconnect (MD).
The detail realization process of the algorithm is
shown in Figure 1. BI sent by stationary node is used to
invite mobile node to join in the communication. More-
over, stationary node would last this action until its reg-
istration form is full. When registration form is full,
compare SNR value of the stationary node to the value of
other nodes which have already connected with mobile
nodes, write the higher value into the form. If BI frame
has existed in the registration form, update registration
form according to the time field in BI frame. MI frame
sent by mobile nodes is to response to BI frame and re-
quest to build up a connection. MR frame is to decide
whether stationary node accepts the MI frame. When
communication link can be established, the stationary
node will send MR frame to mobile node in acknowled-
gement of accepting the connection request. MD frame is
to disconnect the link. After connecting with the statio-
nary node, mobile node continues to listen to this statio-
nary node. When the SNR value of signal which mobile
node receives is lower than the threshold we set before or
there are more appropriate stationary nodes communi-
cating to the mobility, the mobile node sends MD frame
to inform stationary node to interrupt linkage. After sta-
tionary node receives this frame, it deletes information of
mobile node in the registration form.
The disadvantage of EAR algorithm: at first, it could
only be used in the WSN based on TDMA technology;
meanwhile, the network topology could only be cluster
so that the algorithm could ensure the low time delay in
network. Second, energy efficiency of stationary node is
low, because stationary node needs to send extra BI
Figure 1. The communication process of EAR algorithm.

J. J. ZHAO ET AL.
63
frame periodically to find out mobile node. Third, the
algorithm doesn’t propose a suitable sleeping mechanism
based on high redundant network, so it leads to large
number of collision in the high density sensor network
which could result in decreasing the energy efficiency of
node, shorting the life of network and prolonging the
communication latency.
2.2. MS-MAC Algorithm
MS-MAC algorithm adjusts node sleeping time dynami-
cally according to the velocity of mobile node. Its main
idea is based on SMAC protocol. The advantage of the
algorithm is that it proposes a kind of sleeping mechan-
ism which suits for the mobile networks. Similar with
SMAC, MS-MAC algorithm sends SYNC frame to build
up virtual clusters in the period of networking. In every
virtual cluster, all the nodes keep synchronous and the
nodes on the edge of two virtual clusters should follow
synchronous periods of the two virtual clusters. There-
fore, at the beginning of every scheduling mechanism,
nodes send SYNC to build up virtual cluster to keep time
synchronous. When the mobile nodes join in the network,
it must follow time synchronous mechanism of the vir-
tual cluster. So when the mobile node moves from one
virtual cluster to another, it must wait until the next
scheduling mechanism begins, therefore it could know
the time synchronous mechanism of another virtual clus-
ter. Generally, this mechanism increases time delay in
communication. For example, suppose that one schedul-
ing cycle is 2 minutes and nodes stay at wake-up status
in 10 seconds in every 2 minutes, so mobile node at most
needs to wait 1 minute 50 seconds to know synchronous
mechanism of the virtual cluster. MS-MAC algorithm
decreases the time delay by adjusting scheduling me-
chanism dynamically for adapting the velocity of the
mobility. At first, set a value v0 if velocity of mobile
node is higher than 1/4 v0, make scheduling cycle as 1
minute if the velocity of mobile node is higher than 1/2
v0, then make scheduling cycle as 30 seconds, etc.
Though MS-MAC algorithm could use the velocity to
adjust scheduling cycle dynamically, it can’t ensure the
reliable communication between stationary nodes and
mobile nodes; meanwhile, this kind of sleeping mechan-
ism also doesn’t ensure the high rate of coverage and
connection of the whole network.
2.3. LDAS Algorithm
When the node doesn’t know its own coordinate position
information and the angle information of its neighbors,
the sleeping mechanism can use the number of node’s
neighbors to decide whether go to sleep, (the node’s
neighbors defined in the paper are all the nodes in the
circular area whose radius is the node sensing range r),
so that it can cover the most sensing area on the base of
high energy efficiency. The probability formula that the
sleeping node is fully covered by its neighbors is as fol-
lows [3]:


11
1 0.609Pr1 0.609
nn
nAn


 ξ (1)
ξ = 0.5*n*(n-1)*(0.276)n-1. n is the number of the awa-
ken node’s neighbors (n could be selected according to
QoS requirement). In the paper, we suppose that com-
munication radius is equal to sensing radius. So when n =
3, the probability that node can be fully covered is: 0 ≤
Pr(A) ≤ 0.113, when n = 5, then the probability that node
can be fully covered is: 0.312 ≤ Pr(A) ≤ 0.370. Suppose
that n = 3 could satisfy the QoS requirement, meanwhile,
the node finds three more neighbors which are not in
sleeping status, the node considers its sensing rang being
covered by its neighbors and the node could go to sleep-
ing status. In this case, it could happen in following situ-
ations, as Figure (a) shows, the node O1 could be aware
of its three neighbors O2, O3, O4, so it goes to sleeping
status. But when mobile node goes though the shadow
region, any nodes in the network could not be aware of
location of mobile node until the node O1 ends up its
sleeping status. Now we consider the situation that the
smallest sensing area which could be covered by its
neighbors, as Figure (b) shows, suppose that the sensing
radius and communication radius are R. Because central
angle is 120˚, the whole area of shadow region is as fol-
lows:
22
23
34
SR R   (2)
(a)
(b)

J. J. ZHAO ET AL.
64
Because when n = 3, the probability which the node’s
sensing area/communication area fully covered by its
neighbor is among [0, 0.113]. Suppose that R = 10 m, so
we can calculate the area S ≈ 166.032 m2; when n = 5, S ≈
130.833 m2. Therefore, from the above data, it is known
that if only using LDAS sleeping algorithm, there will be
a large number of blind holes in the network, which
could make mobile node disappear.
Although a bigger blind hole occurs in network with a
certain probability, LDAS has its own advantages: 1) the
algorithm is simple that adapts to a large scale network,
the nodes in the network can be random distributed; 2)
the algorithm needn’t know geographical position infor-
mation of neighbor nodes; 3) the algorithm can adjust
dynamically to achieve the requirement of QoS; 4) the
algorithm has a strong resistant ability for channel error
and collision; 5), the blind hole appears randomly and
has a small influence on event monitoring relatively; 6)
the algorithm needn’t keep time synchronous mechanism.
However, LDSA isn’t the mechanism which could real-
ize a reliable point-to-point communication, so it makes
the communication between mobile node and stationary
node unrealized.
3. Design of the DM-MAC Protocol
Combining with the above three kinds of algorithms, we
design a MAC protocol based on distributed network–
DM-MAC. On the premise of ensuring the network cov-
erage, this protocol could realize the reliable communi-
cation between stationary node and mobile node. Mean-
while, the protocol uses the intelligent scheduling me-
chanism proposed in MS-MAC algorithm to decrease the
delay in the communication and to decrease the disap-
pearance probability of mobile node which caused by the
sleeping mechanism.
3.1. Main Design of the Protocol
3.1.1. Netw ork Initialization
Suppose that nodes in the network are distributed ran-
domly and sensing radius of every node r is equal to its
communication radius R. Because the energy of mobile
node can be considered infinitely, so we can increase the
communication radius of the mobile node to 2R by using
power control technology. The initial stage of establish-
ing network, LDSA algorithm will be used to initialize
the sleeping mechanism of every node. The concrete
process is as follows: At first, the stationary node checks
the number of its working neighbors (the number of
working nodes in the sensing range of the node) in every
CT (Checking Time) time. Second, the stationary node
broadcasts its beacon message to show it existing in the
network. The beacon message packet includes two parts:
the nodes’ ID and the number of working neighbors.
When other nodes receive this beacon message, update
related node’s information in the receivers. If the node
doesn’t receive the beacon message send by its neighbors
in a long period, it will make its neighbors turn into
sleeping status. So every node has a neighborhood table.
According to requirement of QoS, we can estimate the
required number of working nodes. If the number of ex-
isting working neighbors is more than required, the node
sends ticket packets to its working neighbors. In working
stage, if the node has enough tickets to turn into sleeping
status, it should also check the number of the working
neighbors again to eliminate the nodes that sleep at the
same time. So if the number of tickets node received is
more than TT (Ticket Threshold), the node turns into
ready-to-off status, and then the node goes to backoff
stage (the backoff time is: random(0, Wmax)). Next, the
node checks for the second time that whether it has
enough working neighbors to make itself turn into sleep-
ing status, if condition is satisfied, the node turns into
sleeping status(off-duty) until passing by a time segment
of ST (Sleep Time). After sleeping a period of ST, the
node goes into on-duty status and clears the values of
tickets.
3.1.2. Communication with St ati ona ry Node and
Mobile Node
When mobile node goes into the network, it broadcasts
BI signal periodically at first (In EAR algorithm, BI is
sent by stationary node which is used to invite the mobile
node to join the communication. In order to decrease the
energy consumption of stationary node and the probabil-
ity of collision, we make mobile nodes to send BI signal).
BI frame includes the velocity of mobile node, the ID of
mobile node and several sensing parameters collected by
mobile node. After the velocity of mobile node (the ve-
locity value can be gained by RF Doppler Shifts [15]) is
received by stationary node, it would be compared with
VT (Velocity Threshold) which is set before the stage of
establishing network, if the two values are different;
changing the ST value defined in the sleeping mechanism
to adjust the node’s sleeping time. In order to make the
stationary node adjust its scheduling mechanism in ad-
vance, we apply power control technology to increase the
mobile node’s transmit power for enlarging its commu-
nication radius. After setting proper scheduling mechan-
ism, mobile node sends BI frame continuously and pe-
riodically, when stationary node receives BI frame, it
sends MI frame to the mobility to ensure that it has re-
ceived the BI frame successfully (in EAR algorithm, MI
frame is sent by mobile node). In the MI frame, it in-
cludes the ID of stationary node and the SNR value of

J. J. ZHAO ET AL.
65
the BI that is received by stationary node. After the mo-
bile node receives MI frame, it accords to SNR value in
the connected communication links to decide whether to
establish the connection, if the SNR value is bigger than
the existing SNR value in the communication links, then
the mobile node interrupts the connected link, establishes
the link to the existing stationary node and then sends
MR frame to establish the connection. If the SNR value
is smaller than the existing SNR value in the communi-
cation links, the mobile node neglects the MI frame and
maintains the existing communication link. In order to
decrease the times of comparison, we set E0 as threshold
of SNR in advance, when SNR value in MI frame is
smaller than E0, the mobile node neglects MI frame di-
rectly. When SNR value in the connected communica-
tion link is smaller than E0 or smaller than any SNR val-
ues in the established communication links, mobile node
must send MD to the stationary nodes to interrupt the
communication link. Notice that, after stationary nodes
have established the communication with the mobile
nodes, even if stationary nodes maybe satisfy the condi-
tion of turning into sleeping status, they can’t go to sleep
until disconnecting the communication with the mobile
nodes.
3.2. The Analysis of DM-MAC Protocol in
Process and Time Sequence
The content of DM-MAC protocol realized in the statio-
nary node and mobile node are not the same, stationary
node mainly implements sleeping mechanism, intelligent
scheduling mechanism and reliable point-to-point com-
munication mechanism, meanwhile, the function of mo-
bile nodes is relative simple, it mainly completes reliable
point-to-point communication mechanism and maintains
SNR table in dynamic neighbor nodes. Here we focus on
the protocol process of stationary nodes in detail. The
basic flow chart is as Figure 2 shows. We suppose that
when nodes send packets, the channel is free. The com-
munication of stationary node includes the communica-
tion between stationary node and mobile node, and the
communication between stationary nodes. Figure 2
shows the communication basic flow on stationary node
and Figure 3 shows the communication basic flow on
mobile node, we apply CSMA/CA mechanism and
S-MAC to realize the communication between the sta-
tionary nodes so that the data which stationary nodes
collect from the surrounding environment and the mobil-
ity information which stationary nodes collect can be
successfully multi-hop to the sink node.
Some nodes are selected to analyze their time sequence.
The topology distribution map is shown in Figure 4. A,
B, C and D are stationary nodes, M is a mobile node, the
sensing area of node B is fully covered by node A, C and
D. Meanwhile, suppose that node B is also fully covered
by its neighbors (In order to describe conveniently, we
don’t consider the case of neighbors of node A, C and D
temporarily). We suppose that when n = 3 can satisfy the
QoS requirement, so if the number of working neighbors
is 3, then node B could go to sleep. When the mobile
node moves into the network, suppose that its communi-
cation radius is r1 = 2R (which is two times bigger than
stationary node’s communication radius), therefore, if
node M moves as Figure 4 shown, node A will listen to
BI frame at first which is sent by M periodically, mean-
while, A adjusts the ST value dynamically by receiving
the velocity in BI frame. According to the SNR in MI
frame, when M continues moving, it will receive more
MI signal, the node will choose the best nodes to estab-
lish communication link (We can set how many number
of stationary nodes should be linked with the mobile
nodes beforehand). The detail time-slot chart of estab-
lishing the link between stationary nodes and mobile
nodes is as Figure 5 shows. We suppose that every mo-
bile node can build communication link with only one
stationary node.
3.3. The Implementation of DM-MAC Protocol
In the design of the whole DM-MAC protocol, we main-
ly implement the reliable point-to-point communication
mechanism between stationary nodes and mobile nodes
with intelligent sleeping mechanism. The sleeping me-
chanism proposed in LDAS algorithm will appear a large
area of blind holes because communication area of node
having been on off-duty status is not fully covered by its
neighbors. The sleeping mechanism proposed in LDAS
algorithm will appear a large area of blind holes because
communication area of node having been on off-duty
status is not fully covered by its neighbors. Thus in
DM-MAC protocol, we enlarge the mobile nodes’ com-
munication radius to make sleeping node foresee the
situation of the mobile nodes in early time, so the nodes
can adjust its own scheduling mechanism after obtaining
the velocity of mobility. The parameters of protocol we
need to set beforehand are shown in Table 1.
The fames involve in the DM-MAC protocol are as
follows: neighborhood table packet, beacon message
frame, ticket frame, BI frame, MI frame, MR frame and
MD frame, their concrete frame structures are shows
from Figure 5 to Figure 11. Since the new protocol is
aimed to the MAC layer, when designing the frame, we
don’t add packaging of physical layer and higher layer in
the frame, such as packaged frame length in physical
layer.
In order to realize the sleeping mechanism which could

J. J. ZHAO ET AL.
66
Figure 2. The protocol flow chart of stationary node.

J. J. ZHAO ET AL.
67
Figure 3. The protocol flow chart of mobile node.
Figure 4. The topology chart of the nodes.
ensure high sensing coverage, every node must maintain
a sheet of neighborhood table. The detail table heading is
followed:
ID is the ID number of the neighbors, Mode is the
current status of the node, NoN is the number of the
neighbors and Update Time is the time when neighbor-
hood updates its content. Meanwhile, in order to com-
pare the SNR value of receiving signal, mobile node also
need maintain a sheet of SNR table. Its table heading is
as follow:
ID is the ID number of transmitting node and SNR
value is the signal-to-noise ratio of receiving signal.

J. J. ZHAO ET AL.
68
Table 1. The input parameters list of MAC layer.
Ticket Threshold (TT) The threshold of the number of tickets node has been received which is to judge the node whether to
go to sleep. It relates to the QoS requirement.
Sleeping Time (ST) The sleeping time of the stationary nodes.
Neighborhood Table Checking (CT) The time interval of updating the nodes’ information in the neighborhood table.
Beacon Message Interval Time (BT) The time interval of node sends beacon message.
Maximum Back-off Time The maximum value of back-off time: Wma x
Velocity Threshold (VT) The threshold is to compare with the velocity of mobile node which is to adjust the sleeping time of
the node dynamically.
SNR Threshold The threshold is to compare the SNR value of received signal which is to decide whether to build the
communication link or not
Figure 5. The time-slot chart of DM-MAC protocol.
Figure 6. The packet structure of neighborhood table.
Figure 7. The packet structure of beacon message.

J. J. ZHAO ET AL.
69
Figure 8. The packet structure of ticket.
Figure 9. The packet structure of BI.
Figure 10. The packet structure of MI.
Figure 11. The packet structure of MR and MD.
Table 2. The table heading of neighborhood table.
ID Mode NoN Update Time (hour:minute:second)
Table 3. The heading of SNR table.
ID SNR value
4. Simulation Results
In order to analyze the performance of DM-MAC proto-
col, we simulate the protocol in PC. Suppose that energy
consumption ratio of transmitting, receiving (idling) and
sleeping is 10:4:0.01. There are 200 stationary nodes
randomly distributed in an area of 150 × 150 m2, sensing
radius and communication radius of the stationary nodes
are 20 m, communication radius of mobile node is 20 m.
Figures 12 and 13 show the case of node distribute and
node sensing/communication coverage. We set ST = 1
min, CT = 10 sec, BT = 15 sec, VT = 3 m/s, TT = 8, the
communication channel is 2405 MHz, when n = 3, QoS
requirement could be satisfied.
4.1. Connection Rate of Mobile Nodes
Under the above condition, comparing with DM-MAC
and LDAS + EAR, we simulate the connection rate of
mobile nodes in the network (the connection rate of mo-
bile nodes = the number of mobile nodes which have
already connected with stationary node/the number of
whole mobile nodes). Suppose that there are 200 statio-
nary nodes randomly distributed in the area, the simulate
results are shows in Figure 14. In the figure, we know
that when the number of mobile node is few, LDAS +
EAR protocol has higher reliability of communication,
but when the number of mobile node increases, packet
loss rate relatively increases and the connection rate of
mobile nodes decreases obviously. When mobile nodes
are added up to 100, nearly 40% mobile nodes disappear.
This is because LDAS + EAR protocol may bring in big
blind holes. Of course collision also brings in disappear-
ing which is the main factor either when the mobile

J. J. ZHAO ET AL.
70
Figure 12. Node distribution.
Figure 13. Node sensing/communication coverage.
Figure 14. The connection rate of mobile nodes.
nodes add, but in the paper we don’t consider this factor
of decreasing connection rate. Although the blind holes
randomly appear, when the number of mobile nodes in-
creases, the probability mobiles appears in the blind
holes will also increases, thus the phenomenon that mo-
bile nodes “disappear” will increase. The other curve in
the figure is the mobile connection rate trend of DM-
MAC, though the connection rate of mobile nodes also
decreases with increasing the number of mobile nodes,
its decreasing rate is obvious slower compared to the
LDAS + EAR protocol. That is because DM-MAC pro-
tocol uses mobile nodes to broadcast BI frame periodi-
cally and stationary nodes to foresee the velocity and the
moving range of the mobiles to adjust sleeping time in-
telligent for decreasing the probability of generating the
blind holes in some degree. Here we consider that the
reliability of network communication is equal to the
connection rate of mobile nodes.
4.2. Sensing Coverage Percentage
Next, we discuss the sensing coverage problem in
DM-MAC (because sensing radius is equal to communi-
cation radius, the problem of sensing coverage is equal to
communication coverage), in the simulation, we bring in
40 mobile nodes, the number of active nodes and cover-
age percentage in the network is shown in Figures 15
and 16. Obviously, as Figure 15 shows, when the Qos
requirement improves (with the increase of n, n is the
number of awaken neighbor nodes), the required active
neighbor nodes need to increase, so it leads to increase
the active nodes in the network. In the Figure 16 we
know that the network coverage of DM-MAC protocol
and LDAS algorithm all achieve 100% with increasing
the number of neighbors required in QoS, but when the
value of n is small, the coverage of DM-MAC is obvious
lager than LDAS algorithm. Because when bringing in
Figure 15. The number of active node.

J. J. ZHAO ET AL.
71
Figure 16. The network coverage degree.
mobile nodes, the scheduling mechanism of stationary
nodes in DM-MAC continues changing (generally the
value of ST becomes small), the number of sleeping
nodes decreases, the coverage area of blind holes de-
creases and the probability of blind holes appearing de-
creases. But when the n value increases, the number of
neighbors also increases, and the probability of blind
holes appearing decreases correspondingly. Though the
simulation test, it is shown that when n = 8, the average
of network coverage in DM-MAC will achieve up to
99.80%, and the average of network coverage in LDAS
will reach 98.40%. Therefore, when n is relative large,
the two kinds of algorithm will approximately realize the
full network being covered.
4.3. Energy Consumption
At last, we discuss the problem of protocol energy con-
sumption. LDAS algorithm only realizes the sleeping
mechanism which doesn’t propose a kind of reliable
point-to-point communication mechanism, so comparing
the energy consumption between LDAS and DM-MAC
is not necessary. Here we give a comparison between
EAR + SMACS protocol, SMACS + EAR and DM-MAC
protocol; we set the occurrence frequency of discrete
event is 0.6 time/m2/sec, n = 3. The result is shown in
Figure 17.
From the above figure, we know that with the simula-
tion time going, the decreasing of energy in EAR +
SMACS is relatively faster, this is because EAR +
SMACS uses the topology of cluster structure, the energy
consumption is large in building cluster stage. Further-
more, in every period of time there will be a process of
choosing cluster head in every cluster, and it applies
TDMA technology to the algorithm which needs precise
time synchronization. All of the above need to consume
large energy. Meanwhile, when the occurrence frequency
Figure 17. The average of node energy consumption.
of discrete event is low, the energy efficiency of EAR +
SMACS is high, but when the frequency increases, the
energy efficiency quickly decreases due to the nodes
awakening frequently. On the other hand, the energy
efficiency of LDAS + EAR and DM-MAC protocol are
relative high, and the two protocols’ energy consumption
are almost the same. This is because the two MAC pro-
tocols is contention-based MAC protocol and they use
period sleeping mechanism simply which save the ener-
gy greatly. In advance analyze, we find that DM-MAC
do better than LDAS + EAR in energy efficiency, this is
because that DM-MAC utilizes dynamic scheduling me-
chanism to adapt to energy saving, also, the most number
of control packets in DM-MAC are sent by mobile nodes
whose energy is considered to be limitless, it decrease the
stationary nodes’ energy consumption greatly. Meanwhile,
the sleeping mechanism of DM-MAC and LDAS + EAR
based on network coverage can basically realize the
energy averaging of nodes in the network, thus prolongs
the network life time.
5. Conclusions
DM-MAC is proposed under the condition of stationary
nodes distributing randomly with a certain amount of
mobile nodes existing in the network, it inherits the ad-
vantage of LDAS algorithm which needn’t to know
neighbors’ geographical position information, dynami-
cally adjusts the network density, has strong anti-inter-
ference ability on channel and collision, blind holes ap-
pears randomly and so on. Meanwhile, DM-MAC com-
bines with the reliable point-to-point communication
which is proposed in the EAR and improves the energy
efficiency of stationary nodes through mobile nodes to
undertake the most transmission tasks, furthermore,
DM-MAC applies power control technology to enlarge
the transmit power of mobile nodes so as to realize the

J. J. ZHAO ET AL.
72
reservation mechanism between mobile nodes and sta-
tionary nodes. In order to decrease the communication
loss rate of mobile nodes and time delay, DM-MAC
combines with the scheduling mechanism in MS-MAC,
which could adjust the sleeping time intelligently.
Generally speaking, DM-MAC protocol is simple to
realize. The energy consumption of every node is nearly
the same. Moreover, the protocol can basically realize
the reliable communication, so it could be applied to the
real time positioning and tracking system. But DM-MAC
protocol doesn’t consider the problem of network con-
nectivity, so there is a problem that the stationary nodes
near the sink node will consume the energy much faster,
thus will make the network paralysis in advance. Our
future work is to solve this problem and analyze the im-
pacts on different values of TT, ST, BT, VT and Wmax for
the performance of network.
6. Acknowledgment
This work is supported by the Scientific Research Fund
of Hebei Province Education Department (2008110) and
Scientific & Technological Research and Development
Projects of Handan in Hebei province (0821103041-2).
7. References
[1] V. Rajendran, K. Obraczka and J. Garcia-Luna-Aceves,
“Energy-Efficient, Collision-Free Medium Access Con-
trol for Wireless Sensor Networks,” Proceedings of the
ACM Sensor Systems 2003, Los Angeles, 2003, pp. 181-
192.
[2] A. El-Hoiydi and J.-D. Decotignie, “WiseMAC: An Ultra
Low Power MAC Protocol for Multi-hop Wireless Sensor
Networks,” Algorithmic Aspects of Wireless Sensor Net-
works First International Workshop, Vol. 3121, 2004, pp.
18-31.
[3] K. Wu, Y. Gao, F. Li and Y. Xiao, “Lightweight Dep-
loyment-Aware Scheduling for Wireless Sensor Net-
works,” ACM/Kluwer Mobile Networks and Applications
(MONET) Special Issue on Energy Constraints and Life-
time Performance in Wireless Sensor Networks, Vol. 10,
No. 6, 2005, pp. 837-852.
[4] K. Sourabi, J. Gao, V. Ailawadni and G. J. Pottie, “Pro-
tocols for Self-Organization of a Wireless Sensor Net-
work,” IEEE Personal Communications, Vol. 7, No. 5,
2000, pp. 16-27. doi:10.1109/98.878532
[5] H. Pham, and S. Jha, “An adaptive Mobility-Aware MAC
Protocol for Sensor Networks (MS-MAC),” IEEE Inter-
national Conference on Mobile Ad-hoc and Sensor Sys-
tems, 25-27 October 2004, pp. 558-560.
[6] W. Ye, J. Heidemann and D. Estrin, “Medium Access
Control with Coordinated Adaptive Sleeping for Wireless
Sensor Networks,” IEEE/ACM Transactions on Net-
working, Vol. 12, No. 3, 2004, pp. 493-506. doi:10.1109/
TNET.2004.828953
[7] W. Ye, J. Heidemann and D. Estrin, “An Energy-Effi-
cient MAC Protocol for Wireless Sensor Networks,”
INFOCOM 2002, pp. 1567-1576.
[8] B. Kusy, A. Ledeczi and X. Koutsoukos, “Tracking Mo-
bile Nodes Using RF Doppler Shifts,” ACM Sensor Sys-
tems, 2007, pp. 29-42.
[9] L. Wang and Y. Xiao, “A Survey of Energy-Efficient
Scheduling Mechanisms in Sensor Networks,” Mobile
Networks and Applications, Vol. 11, No. 5, 2006, pp.
723-740. doi:10.1007/s11036-006-7798-5
[10] P. Raviraj, H. Sharif, M. Hempel and S. Ci, “MOB-
MAC—An Energy Efficient and Low Latency MAC for
Mobile Wireless Sensor Networks,” 2005, pp. 370-375.
[11] T. V. Dam and K. Langendo, “An Adaptive Energy- Ef-
ficient MAC Protocol for Wireless Sensor Network,” The
First ACM Conference on Embedded Networked Sensor
Systems, 2003, pp. 171-180.
[12] M. Ali, T. Suleman and Z. Uzmi, “MMAC: A Mobili-
ty-Adaptive, Collision-Free MAC Protocol for Wireless
Sensor Networks,” Proceedings of 24th IEEE Perfor-
mance, Computing, and Communications Conference,
2005, pp. 401-407.
[13] D. Tian, N. D. Georganas, “A Coverage-Preserving Node
Scheduling Scheme for Large Wireless Sensor Net-
works,” Proceedings of the 1st ACM International
Workshop on Wireless Sensor Networks and Applications,
2002, pp. 32-41. doi:10.1145/570738.570744
[14] A. L. Henrique, F. Daniel, L. dos Aldri, S. J. Marcos, and
A. A. Ferreira, “A Taxonomy for Medium Access Con-
trol Protocols in Wireless Sensor Networks,” Annales des
Télécommunications, Vol. 60, No.1, 2005, pp. 944-969.
[15] H. Yousefi’zadeh, H. Jafarkhani and J. Kazemitabar, “A
Study of Connectivity in MIMO Fading Ad-Hoc Net-
works,” Journal of Communication and Networks, Vol.
11, 2009, pp. 47-56.