Int. J. Communications, Network and System Sciences, 2010, 3, 280-288
doi:10.4236/ijcns.2010.33036 blished Online March 2010 (
Copyright © 2010 SciRes. IJCNS
Spectrum Efficiency Improvement Based on the Cognitive
Radio Management
Jamal Raiyn
QRC-Qasemi Research Center, Alqasemi College, Baka Alqarbiah, Israel
Received November 3, 2009; revised December 5, 2009; accepted January 19, 2010
Interference and delay are considered as the major reasons limiting the capacity and increasing the new call
blocking probability in cellular system. In this paper we introduce a novel strategy based on cognitive radio.
Cognitive radio is defined as a radio or system that senses its environment and can dynamically and autono-
mously change its transmitter parameters based on interaction with the environment in which it operates,
such as maximize throughput and reduce interference. The goal of the use of cognitive radio is to improve
the spectrum efficiency in cellular system. Spectrum management based on radio cognitive plays thereby an
important role to increase the capacity of the radio systems and spectrum utilization, especially in the context
of open spectrum.
Keywords: Cognitive Radio, Handover, Social Agent
1. Introduction
In this paper we introduce cognitive radio approach that
is expected to perform more significant role in the view
of efficient utilization of the spectrum resources in the
future wireless communication networks. The spectrum
utilization efficiency is defined as the ratio of informa-
tion transferred to the amount of spectrum utilization.
Our approach is reactive approach, in that it enables, via
negotiation, learning, reasoning, prediction, active sense,
identification, changes in the base station’s parameter to
meet the new services requirements in modern wireless
networks and future challenges in cellular systems. A
major challenge with cognitive radio approach is to be
done in near real-time and to keep up with an ever
changing RF environment without overly computation-
ally complex. Various resource allocation strategies are
proposed to optimize the resource allocation in cellular
system by reducing the call blocking probability [2,3,
5,6,8] in cellular systems. The call blocking probability
is often measured in terms of two blocking probabilities,
the arriving call blocking probability, and the handover
blocking probability. Analyses and studies in [9–11]
show that the call blocking probability in handover is
caused by two main parameters, interference and delay.
Interference leads to missed and blocked calls due to
errors in the digital signaling. Between transmitter (Base
Station, BS) and receiver (Mobile Station, MS), the cha-
nnel is modeled by several key parameters. These pa-
rameters vary significantly with the environment (urban,
rural, mountains).There are different type of interference
that when not minimized, decreases the ratio of carrier to
interference power at the periphery of the cells, causing
diminished system capacity, more handover [1,4], and
more dropped calls. To reduce the handover blocking
probability in cellular systems has been proposed various
schemes as, prioritized handover schemes and handover
with queueing [4]. In some application fields like real-
time communication and industrial automation is needed
to ensure a seamless and lossless handoff. Which means
the handover latency should be zero. For efficient hand-
over management, Handover is a basic mobile network
capability for dynamic support of terminal migration.
Handover Management is the process of initiating and
ensuring a seamless and lossless handover of a mobile
terminal from the region covered by one base station to
another base station. In this paper we consider the hand-
over call blocking probability in cellular systems.
The paper is organized as follows: Section 2 intro-
duces the system model and the social agent strategies.
Section 3 discuss and analyze the handover call blocking
probability based on simulation results and Section 4
concludes the paper.
2. Problem Des cription
Over the last two decades, the demand for mobile host
J. RAIYN 281
and multimedia services increased rapidly. One of the
biggest challenges in supporting multimedia applications
in cellular systems is to fulfill the mobile user demand
and satisfy his preferences under the constraint of the
limited radio bandwidth, and to utilize the limited spec-
trum availability to meet the increasing demand for mo-
bile service. Some of the most often used methods to
increase the spectral efficiency are resource allocation
schemes. Various channel allocation schemes have been
introduced to provide Quality of Service (QoS) [14,15,
16] and efficient channel utilization in cellular networks.
There are many parameters to measure the QoS of a
network. These include throughput, latency, service
availability etc. The blocking probability is one of the
most important QoS parameters. Since users are mobile,
the QoS of wireless networks are often measured in
terms of two probabilities: the new arriving call blocking
probability and the handoff blocking probability. Hence,
this paper deals with the main issue: How to allocate
resources (e.g. frequency channels) to radio ports of a
wireless system (e.g. cells in a cellular mobile network)
that can improve the traffic performance of network (e.g.
lower blocking probability in voice networks, lower la-
tency in data networks etc.)?
In the modern data communication systems, we con-
sider the power transmission as the resources. An in-
crease of transmission power enables the increase of data
rate and subsequently shorter transmission time. On the
other hand, an increased transmission power causes more
interference in neighboring cells. One of the physical
measures of RF channel quality is the carrier-to-inter-
ference or CIR. This ratio is logarithmically proportional
to the signal quality enjoyed by the receiver of the signal.
The larger the C/I ratio, the better the channel quality is.
If the measured C/I falls below a certain level, CIRmin,
which depends on system type and operator requirements,
the mobile should be in the coverage region of another
cell and a call handoff should be performed. The interior
of the cell should provide C/I ratio which exceed this
level, unless the mobile is located in an RF coverage
“hole.” In general, for a feasible solution for the resource
managements in the cellular systems must be satisfied
four requirements as follows:
Coverage requirement: each of the MS locations mu-
st be able to connect at least to one base station and the
received signal strength must exceed a given level Pmin.
Transmitter power requirements: the transmitted
power of the mobile stations in the system must not
exceed a maximum level Pmax.
Downlink SIR requirement: each downlink SIR
must exceed a threshold of SIRth.
Uplink SIR’s requirement: each uplink SIR must
exceed a threshold of SIRth.
In this case, we consider the problem of minimizing
total transmitter power. To minimize the total transmit
power, subject to the constraint that each transmitter/
receiver attain a maximum allowed outage probability
i.e., a minimum allowed quality of service and subject to
limits on the individual transmitter power, we formulate
the problem as follows:
minimize n
Subject to , i = 1, …, n
min max
Here, and are the minimum and maxi-
mum transmitter power for the transmitter i. To handle
the resource allocation problem in cellular systems we
introduce the cognitive radio scheme. For the channel
allocation problem, the main goal is to bring more dis-
tribution of control. In cognitive radio scheme, each cell
(base station) is managed by an agent. The resource in
the system will be distributed and allocated through in-
teraction between agents and limited to the domain of the
agents. In other words, the channel allocation problem
can be restricted to one domain, so that the agents can be
placed in the base stations.
3. Power Control Based on CR
Each cell i can acquire amount i while the level in the
cell is denoted by hi (the demand of channel in each cell
depends on the arrival call in the cell). We assume, that
the mass flow of resource in the cell i is proportional to
the value position ui (call arrival rate). And we consider
that the resources in the system are limited. The vector of
the required set points of the levels, (to meet the call ar-
rival rate), are given by hs. We further assume that the
equilibrium point (u0, hs) is known (channel allocation).
The control task is the regulation of disturbances (plus
the interference) of the resources in the cells. Each cell is
considered as a single subsystem and associated with an
agent. The input variable of each agent i is the deviation
hi(k) of the resource amount.
hkh h
 (1)
The output variable is the value position
() ()
ukuk u
 (2)
which is the sum of the deviation ui(k) and the equilib-
rium position u0. Since the equilibrium position is known,
the agent only has to compute the output variable ui(k).
Since the amount of the resources that flows into the cell
is limited, the absolute Value |ui(k)| is considered as the
resource variables. Now each agent can either play the
role of the consumer or the producer. A positive devia-
tion hi(k) means that the cell “needs” some more re-
source to reach the required level (i.e. to allocate all
hkh h
 > 0. i.e., hi > hs (3)
opyright © 2010 SciRes. IJCNS
Thus, the related agent can be declared as a consumer,
which has to acquire a certain amount of the resource.
This demand of consumer i at instance k is denoted by
rDi(k). The goal of the consumer is to maximize his util-
ity, (i.e. by reducing the call blocking probability). JCi(k)
() ()() ()
Jk hkrkIk
 (4)
The relative utility is a composition of two terms. The
first term takes into account that the utility of the cell
from getting some resource increases with increasing
deviation hi(k). The second term denotes the con-
sumer’s expenditure (allocation) with respect to his in-
terference I(k). Its utility decreases with increasing ex-
penditures. The minimization of the call blocking prob-
ability JCi(k) leads to the consumer’s demand:
Di i
rk hk
  (5)
The demand is a function of the interference. An in-
creasing of the interference leads to a decreasing demand.
In the case of a negative deviation hi(k), the cell con-
tains too much free resources. Hence, the related agent is
considered as a producer which wants to gift this re-
source to the consumers with the goal to maximize sys-
tem capacity/efficiency and its profit Jpi(k), denoted by
() |()|*()(
Jk hkIkr k
rk )
The relative profit contain two terms; The first term
takes into account that the utility of the producer in-
creases with his income and with increasing deviation
|hi(k)|. The second term is production costs that are as-
sumed to be proportional to the supply. The maximiza-
tion of Jpi(k) leads to the supply of producer agent i:
() |()|()
Si i
rkhk Ik (7)
One problem occurs: If all agents are producers or all
agents are consumers. However, there is always nominal
channels/resources in the cells (nominal channel can be
used only from own cell). Therefore, obviously have a
permanent producer, which is independent of the amount
of resource (swapping method). Assuming n coupled
cells, the actual number of producers at the instance k is
m while the actual number of consumers is q with n = m
+ q. Each consumer agents computes his demand while
each producer agents computes his supply. All functions
are transmitted to an auctioneer agent using the commu-
nication network (the auctioneer could be the distributor
agent, (D). The task of the auctioneer is the computation
of the equilibrium using the constraint that the sum of all
demands has to equal the sum of all supplies:
()() ()()
DpDi SpSi
rk rkrkrk
Hence the output variables ui(k) are
()() ()
uk hk
, i consumer.
()() ()
ukhk Ik
 , i producer.
• Power control
The power control can raise the network capacity.
Some power control algorithms [17,18,19,20] based on
the idea of balancing the SIR of all radio links have been
introduced, but the final SIR achieved by those algo-
rithms may be unsatisfactory for some of the links. Some
calls must be dropped in order to keep the SIR of other
calls higher than the predefined threshold value. Obvi-
ously, the efficiency of radio resource management is
dependent on the channel assignment and the power con-
trol. The combination of DCA and power control to ob-
tain some substantial capacity gains has been reported in
[22], however, because no channel pre-selection is done
before the channel probing procedure, inadvertent drop-
ping of calls caused by originating calls can occur so
often that all unsuccessful (blocked or dropped) calls are
unintentionally dropped calls and not blocked calls. In
addition, an exhaustive search and too frequent intra-cell
handoff access will decrease the system capacity and
make the algorithms difficult to implement in real net-
works. Here, a cognitive radio algorithm with power
control is proposed. The power that is transmitted both
from the mobile equipment and from the base station has
a far-reaching effect on efficient usage of the spectrum.
Power control is an essential feature in mobile networks,
in both uplink and downlink. When a mobile transmits
high power, there is enough margin in the critical uplink
direction. But it can cause interference to other sub-
scriber connections. The power of the signal transmitted
by the base station antenna should be kept to a level
above the required threshold without causing interfer-
ence to the mobiles. Mobile stations thus have a feature
such that their power of transmission can be controlled.
This feature is generally controlled by the BSS. This
control is based on an algorithm that computes the power
received by the base station and, based on its assessment,
it increases or decreases the power transmitted by the
mobile station. The signal power at the ith receiver is
given by GiiFiiPk [21], and the total interference power is
given byikik k
The SIR of the ith receiver (or transmitter) is given by
iiii k
ikik k
Copyright © 2010 SciRes. IJCNS
J. RAIYN 283
iiii k
We assume that the QoS requested is provided when
the SIR exceeds a given threshold SIRth. The outage
probability of the ith receiver/transmitter pair is given by
Pr ()th
= (11)
Pr th
iiiiiikik k
The outage probability Oi can be interpreted as the
fraction of time the ith transmitter/receiver pair experi-
ences an outage due to fading. Note that the in our ex-
pression for Oi, we take into account statistical variation
of both received signal power and received interference
power. We now consider the market method to regulate
the SIR. By ignoring all statistical variation of both
signal and noise power, the signal power at the ith re-
ceiver is then GiiPi and the interference power at the re-
ceiver is given by . Then the SIR at the ith re-
ceiver is given as
ik k
mii i
ik k
We interpret as follows: this is what the sig-
nal-to-interference of the ith transmitter/receiver pair
would be, if the fading state of the system where F1= …
= Fn = 1. We also define
min min
ik k
which is the minimum SIR of the system over all trans-
mitter/receiver pairs. Like the outage probability O, the
SIRm gives a figure of merit for the system and power
allocation. We define the market regulation/control met-
hod of the system and power allocation as the ratio of the
market control SIR to the signal-to-interference reception
ii i
th th
iik k
 (14)
There is a relation between MCA and O: when MCA
is large (which means that the SIR, ignoring statistical
variation, is well above the minimum required for recep-
tion), we should have small O. Let
denote the channel variation. i
will be estimated and
predicted in the proposed power control scheme.
presents the received interference. is
the link gain from mobile station k to base station i.
Suppose, that the highest transmitted power allowed is
Pmax and that the lowest transmitted power allowed is
Pmin. The social agent that SNR oriented uses the fol-
lowing technique to regulate the SNR between the cells.
The proposed technique operates in the following way:
• For any cell, two tiers of cells are considered as in-
terfering cells. The channel state information (allocating
or releasing) of each cell is locally exchanged to its in-
terfering cells. Every cell maintains a list of the cost for
all channels. The cost function is used to decide the cost
of a channel. The cost of a channel in a cell is updated
(increased or decreased) in real time if a co-channel call
is accepted or terminated (dropping and departure) in one
of the cell’s interfering cells.
• When a call arrives in a cell, the free channel with
highest priority (lowest cost) is chosen for call setup and
the call power probing process is activated. The proce-
dures of the power probing for a new (or handoff) call
1) Assigning the minimum transmitted power Pmin to
the new call p.
2) Measuring the SIR value γp of the call.
3) If γp < γ, adjusting the power of the call and going
back to step 2; if γp γ, and Pmin Pp(k) Pmax, this call
is admitted into service with this power and the call
power probing process is ended.
4) If a power cannot be found in the range of [Pmin,
Pmax] with which the SIR value γp γ, or the probing
iteration number is larger than a pre-assigned value, the
probing is moved to the next highest priority channel.
Actually, an exhaustive searching is not allowed in a
system. Hence, we prescribe that if four channels have
been evaluated, but the SIR requirement is still not satis-
fied, the call is blocked.
5) If a call is in service, the power control algorithm is
used to maintain its quality. Each base station monitors
its own served calls at some time interval. We assume
that all base stations are synchronized (actually the algo-
rithm works asynchronously either). When the SIR of a
call falls below the target value, the power control pro-
cedure is requested. However, if the maximum transmit-
ted power is requested or the number of iterations of
power level adjustment is larger than the allowed value,
but the SIR is still below a specified value (e.g., the call
dropping threshold value), the handoff procedure is re-
quested. The “call set-up” procedure will begin to search
for a channel for handoff. If a channel is found, the call
is moved to this channel. Otherwise, the call is dropped.
4. System Model
A cellular network consists of an array of cells. We par-
opyright © 2010 SciRes. IJCNS
tition the cellular network to clusters. For each cell we
set a social agent, as depicted in Figure 1. By using radio
cognitive approach we aim to achieve an optimal net-
work capacity, minimizing interference to other signals
and to reduce messages complexity and channel acquisi-
tion delay that are considered the main reasons to block
the new calls. A radio cognitive approach may be able to
sense the current spectral environment, and have infor-
mation of past transmitted and received packets along
with the power, bandwidth, and modulation. By consid-
ering all this, it makes better decisions about how to best
optimize for some overall goal. Under heavy traffic load,
and if a vacant channel is not found, the social agent
then tries to obtain an exclusive channel by optimization
of channel distribution based on iterative swapping sche-
me [23]. In which the social agent changes the channel
distribution anew. We partition the set of channels into
active and passive channels. The active channels are de-
fined as the channels, which can be used by own cell.
Furthermore the active channels can be simultaneously
used in different cells without any interference if they are
a minimum reuse distance (Dmin) apart from each other
[23]. The passive channels are defined as the channels,
which can be used by neighbor cells. Furthermore, the
free channel will be expressed as 1 and the busy channel
will be expressed as 0. The set of channels are classified
and assigned uniform to cells in real-time.
4.1. Social Agent Negotiation Strategy
Compared to the traditional negotiation strategy that of-
fers high messages complexity, we are interested in ap-
Figure 1. System model.
plications where negotiation between social agents ser-
ves to solve the resource allocation problem in cellular
systems. Furthermore we anticipate concern the feasibil-
ity of reaching an allocation of resources that is optimal
from a social point of view. Social agents often need to
interact in order to improve their performance. One type
of interaction that is gaining an increasing interest is dy-
namic negotiation. The goal of negotiation is the maxi-
mization of the utility of a future decision. In distributed
dynamic environment, each cell has an objective that
specifies its intention to acquire a free channel for call
establishment. That objective should be achieved in a
certain amount of time, specified by a deadline. Negotia-
tion stops when this deadline is reached.
4.2. Social Agent Decision Strategy
In this phase the agent deals with handover request as
illustrates in Figure 2. The social agent’s decision about
the handover process is focused on quality of service
requirements (e.g. signal power, signal-to-noise ratio and
delay). The signal-to-noise ratio (SNR or S/R) defined as
the ratio of a signal power to the noise power corrupting
the signal. The social agent estimates the SNR and then
determines to carry out the handover or not. Furthermore
the agent collects information about the adjacent cells.
Based on of the collected information from the adjacent
cells, the agent determines the next handover cell. The
handover based on SNR can be divided into two main
1) The first scenario is based on received SNR from
the base station only. This method decides handoff when
the SNR from current station is smaller than another sta-
tion. This kind of method is simple but will take place
repeated or unnecessary handoff.
2) The second scenario is based on relative SNR with
threshold. In this approach, handoff is initialed when the
average SNR falls below a certain threshold value. This
method can avoid unnecessary handoff when the current
station signal is still satisfactory.
area rejected
accepted /
cell-A cell-B
Figure 2. Blocking probability area.
Copyright © 2010 SciRes. IJCNS
J. RAIYN 285
4.3. Social Agent Reasoning Strategy
The radio approach draws its decision based on employ-
ing a cost function [24]. The resource agent allocates to a
request, the channel of which the cost function is mini-
mal. The cost function basically computes the interfer-
ence level. The cost functions can be collectively ex-
pressed in a general expression:
where k
is the channel interference cost unit for the
k-th channel, c
denotes the set of co-channel interfer-
ence cells related to cell C. denotes the binary
status of
which signifies that
0, channel causes no interf. in cell
, channel causes interf. in cell
C1k i
ki is used to reflect the interference between the in-
terfering cell i and cell C. Therefore, any available
channel having minimum value of k
is to be allocated
to a new call arising in cell C. To obtain an optimal rea-
soning, the social agents consider aggregation rules that
enable the social agents to change the currently held
reasoning upon acquiring new information. Information
and actions that affect the social agent decision in hand-
over process are described below.
Handover process starts if the mobile station re-
ceives a weak SNR.
Mobile station moves to neighbor cell that have the
highest number of free channel to ensure that the call is
not blocked.
Chooses channel with the lowest interference from
the set of free channels.
() argmink
channel kI
Using iterative swapping scheme to avoid high in-
t = 0
FOR i = 1 to k DO r
i(t) = choose (V)
FOR i = 1 to k DO C
i =
vtrdx i
FOR i = 1 to k DO
ri(t) = minimize (interf(Ci))
ri : d(ri(t), ri(t-1)) <
, t > tmax
RETURN ({r1(t), ... , rk(t)})
4.4. Social Agent Beliefs
The new call in cell is blocked when there are no more
free channels in the cell or the QoS requested cannot be
provided as the SIR is under a given threshold SIRtgt. By
computing the call blocking probability in handover
process, we consider four social agent decision scenarios
that are described below.
1) Probability (Approve, SNR > SNRtgt)
2) Probability(Reject, SNR < SNRtgt)
3) Probability(Approve, SNR < SNRtgt)
4) Probability(Reject, SNR > SNRtgt)
4.5. Interference Area Identification
In a mobile system, since the mobile stations can move
between cells, the number of mobile stations within a
cell at a given time can never be known exactly in ad-
vance. However, we can estimate roughly the location of
the mobile stations. To avoid the call blocking by hand-
over, it is important to identify the interference area. The
interference area can be considered as the node of deci-
sion to hand over to a new cell. Denote Pt as the trans-
mitter power of the base station, G the antenna gain, d
the distance between the transmitter and receiver and N0
as the thermal noise power. Generally, then, the received
signal-to-noise ratio (SNR) at the k-th user in cell-1 is
given by
11 1
1( )2()()
()( )( )
,,..., k
dd jd
PG dPG dPG d
 
 
for j = 1,…,k. The signal-to-noise ratio (SNR or S/R) k
defined as the ratio of a signal power to the noise power
corrupting the signal. Hence, following Markovian
analysis, the social agent can predict the average block-
ing probability of user k being served by the base station
based on calculation of the interference area for cell i,
which is greater than a given target value for cell i, tgt
This blocking probability is:
1 0
4.6. Active Sense Environment
In sense environment the radio cognitive approach takes
in consideration the following parameters, propagation
model, traffic model and amount of the information. This
represents the maximum amount of information that can
be conveyed through a communications channel. From
an information theoretic perspective, a communications
channel is responsible for passing data between two
points, and will likely add some sort of noise to be origi-
nal signal. In other words, the original signal reception is
opyright © 2010 SciRes. IJCNS
possible only when the relation of energy per bit Eb to
noise spectral density N0 is appropriate. Low value of
Eb/N0 will cause the receiver to be unable to decode the
received signal, while a high value of the energy per bit
in relation to noise will be perceived as interference for
other users of the same radio channel. For example, for
CDMA systems, the bit energy-to interference-and-noi-
se-spectral density (Eb/N0) is SINR multiplied by the
number of information bits modulated by the spreading
code, whereas the carrier-to-interference-and-noise-pow-
er ratio (CIR or C/I) is equal to (Eb/N0) divided by the
length of the spreading code. The ratio Eb/N0 is given by
EN GCI (19)
where Gp is the processing gain, and the ratio C/I of the
user is given by
To calculate the received signal-to-noise ratio (SNR)
at the kth user is given by
 
PG r
For our purposes we focus on the Shannon Theorem,
which states for additive white Gaussian noise (AWGN)
channel. The channel capacity is given by
log 1k
is the bit error rate.
5. Simulation Results
The radio cells treat licensed users, other unlicensed ra-
dio networks, interference, and noise all as interference
affecting the signal-to-interference ratio (SIR). Higher
interference yields lower SIR, which means lower capac-
ity is achievable for a particular signal bandwidth and
interference in the radio channel and reduces the quality
of the transmission. There are different quantities that
measure the quality as signal-to-interference ratio (SIR)
and the bit-error rate (BER). SIR, referred to also as sig-
nal-to-interference-and-noise ratio (SIR) to emphasize
the presence of background noise, is the ratio between
the power of the desired signal and the power of the in-
terference (plus noise). In Figure 3, two different mobile
radio systems are illustrated. Mobile station MSA is lo-
cated at the cell boundaries of system A, however very
close to base station BS-B, due different reasons like
receiving a weak SNR (finding the mobile station in rural
area) or due interference that may occur at base station
BSA from base station BS-B. To maintain a reliable con-
nection between the user and the base station, the SIR at
the receiver should be no less than some threshold that
corresponds to QoS requirement such as the bit error rate.
Figure 4 shows that the received signal to interference
ratio varies greatly over the duration of the simulation.
Figure 5 describes the average received power by the
Et hernet
decision maker
Figure 3. Traffic model.
050 100150 200 25
Time (s)
Rec eived SNR (dB)
Figure 4. The received SNR.
050100 150 200 250300 350 400
Distance (meters)
Average r eceived po w er (d Bm)
Figure 5. The received power vs. distance.
Copyright © 2010 SciRes. IJCNS
J. RAIYN 287
users. Figure 6 describes the social agent strategy that
may agree on a deal to exchange some of the resources
they currently hold, in order to increase the social agent
utility. Figure 7 describes the bandwidth efficiency. Fig-
ure 8 describes the blocking probability related to the
200 300
Distance (m)
Average recieved power (dBm)
Number of the cells
Figure 6. Resource distri bution.
12 14
C (bps/Hz)
Figure 7. Bandwidth efficiency.
Blocking probability
Figure 8. Blocking probability.
6. Conclusions
In this paper we have presented the principles of handoff
procedures and described some of the procedures used in
various types of systems. Furthermore we have proposed
a radio cognitive for handover management to reduce the
interference which is sourced by channel acquisition in
cellular system. In general, there are different reasons
that caused the interference in cellular system, such as
power which is also an important resource. Allocation of
power in the proper channels can increase capacity and
avoid interference.
7. References
[1] N. D. Tripathi, et al., “Handoff in cellular systems,” IEEE
Personal Communications, December 1998.
[2] B. Friedlander and S. Scherzer, “Beamforming versus
transmit diversity in the downlink of a cellular commu-
nication,” IEEE Transactions on Vehicular Technology,
Vol. 53, No. 53, July 2004.
[3] C. W. Leong, W. Zhuang, Y. Cheng, and L. Wang, “Op-
timal resource allocation and adaptive call admission con-
trol for voice/data integrated cellular networks,” IEEE
Transactions on Vehicular Technology, Vol. 55, No. 2,
March 2006.
[4] L. Cong, et al., “Performance analysis of dynamic chan-
nel assignment algorithms in cellular mobile systems with
hand-off,” International Journal of Communication Sys-
tems, July 2002.
[5] S. S. Manvi and P. Venkataram, “Agent based subsystem
for multimedia communications,” IEE Proceedings Soft-
ware, Vol. 153, No.1, February 2006.
[6] S. Anand, A. Sridharan, and K. N. Sivarajan, “Perform-
ance analysis of channlized cellular systems with dy-
namic channel allocation,” IEEE Transactions on Ve-
hicular Technology, Vol. 52, No. 4, July 2003.
[7] T. Ren and R. J. La, “Downlink beamformig algorithms
with inter-cell interference in cellular networks,” IEEE
Transactions on Wireless Communications, Vol. 5, No.10,
October 2006.
[8] M. Zafer and E. Modiano, “Blocking probability and
channel assignment in wireless networks,” IEEE Trans-
actions on Wireless Communications, Vol. 5, No. 4, April
[9] S. Gupta and P. Srimani, “Distributed dynamic channel
allocation in mobile networks: Combining search and
update,” The Journal Supercomputing, Vol. 17, No. 1,
August 2000.
[10] S. Kandukuri and S. Boyd, “Optimal power control in
interference limited fading wireless channels with outage
probability specifications,” IEEE Transactions on Wire-
less Communications, Vol. 1, No. 1, January 2002.
[11] C. J. Chen and L. W. Wang, “Impacts of radio channel
opyright © 2010 SciRes. IJCNS
Copyright © 2010 SciRes. IJCNS
characteristics, heterogeneous traffic intensity, and near-
far effect on rate adaptive scheduling algorithms,” IEEE
Transactions on Vehicular Technology, Vol. 55, No. 5,
September 2006.
[12] J. Raiyn, et al., “Using adaptive agent-based on iterative
distributed swapping prediction for interference reduction
in cellular systems, international symposium on per-
formance evaluation of computer and telecommunication
systems,” Endinburgh, U.K., 1618 June 2008.
[13] M. Bublin, M. Kongegger, and P. Slanina, “A cost-func-
tion-based dynamic channel allocation and its limits,”
IEEE Transactions on Vehicular Technology, Vol. 56, No.
4, July 2007.
[14] M. S. Do, Y. Park, and J. Y. Lee, “Channel assignment
with QoS guarantees for a multiclass multicode CDMA
system,” IEEE Transactions on Vehicular Technology,
Vol. 51, pp. 935948, September 2002.
[15] A. Seth, H. M. Gupta, and K. Momaya, “Quality of ser-
vice parameters in cellular mobile communication,” In-
ternational Journal of Mobile Communications, Vol. 5,
No. 1, pp. 6593, 2007.
[16] S. Zhao, Z. Xiong, and X. Wang, “Optimal resource allo-
cation for wireless video over CDMA networks,” IEEE
Transaction on Mobile Computing, Vol. 4, No. 1, pp.
441424, JanuaryFebruary 2005.
[17] G. J. Foschini and Z. Miljanic, “A simple distribute
autonomous power control algorithm and its conver-
gence,” IEEE Transactions on Vehicular Technology,
Vol. 42, No. 4, pp. 641646, November 1993.
[18] S. A. Grandhi, J. Zander, and R. Yates, “Constrained
power control,” Wireless Personal Communications, Vol.
1, No. 4, pp. 257270, 1995.
[19] B. C. Jung, Y. J. Hong, D. K. Sung, and S. Y. Chung,
“Fixed power allocation with nulling for TDD-based cel-
lular uplink,” IEEE Communications Letters, Vol. 12, No.
4, April 2008.
[20] X. Li and D. Wu, “Power control and channel allocation
for real time applications in cellular networks,” Wireless
Communications and Mobile Computing, Vol. 8, No. 6,
pp. 705713, 2007.
[21] J. L. Eaves and E. K. Reely, “Principles of modern ra-
dar,” Van Nostrand Reinhold Company, New York,
[22] G. J. Foschini and Z. Miljanic, “Distributed autonomous
wireless channel assignment algorithm with power con-
trol,” IEEE Transactions on Vehicular Technology, Vol.
44, No. 3, pp. 420429, August 1995.
[23] D. Beckmann and U. Killat, “A new strategy for the ap-
plication of genetic algorithms to channel assignment
problem,” IEEE Transactions on Vehicular Technology,
Vol. 48, No. 4, pp. 12611269, 1999.
[24] M. Bergounioux and K. Kunisch, “On the structure of
Lagrange multipliers for state-constrained optimal control
problems,” Optimization and Control of Distributed Sys-
tems, Vol. 48, No. 34, pp. 169176, March 2003.
[25] S. Hykin, J. D. Thomson, and H. J. Reed, “Spectrum
sensing for cognitive radio,” Proceedings of the IEEE,
Vol. 97, No. 5, pp. 849877, May 2009.