Content uploaded by Ebenezer Esenogho
Author content
All content in this area was uploaded by Ebenezer Esenogho on Dec 31, 2014
Content may be subject to copyright.
Impact of Primary Users on Secondary Users Channels in a Centralized
Cognitive Radio networks
ESENOGHO EBENEZER, TOM WALINGO AND FAMBIRAI TAKAWIRA
Discipline of Electrical, Electronics and Computer Engineering
University of KwaZulu-Natal, 4041, Durban
King George Avenue
South Africa
213573292@stu.ukzn.ac.za, tom@ukzn,ac.za
Abstract: - Cognitive radio network is a research paradigm in solving the problem of TV white space spread
across spectrum. However, this underutilized resources that is facing scarcity is being used by unlicensed
secondary users opportunistically when the primary users are idle i.e., when the channels are not occupied by
primary users (PUs). Thus, maximizing the spectrum and vacating the channels before the primary users arrives
since they have higher priority. If primary user arrives and the secondary user is till transmitting packets, by the
principle of overlay, the SU is forcefully terminated or buffered and if the PU arrives and enough channel exist
how does it affect the SU performance. This paper investigates the impact of the primary users on the
secondary channel in a cognitive radio network in terms of the throughput, average service time, delay etc. The
occupancy of the primary channel is modelled as a discrete-time two Markov chain also, a simplified analytical
model is presented to obtain the performance of an Opportunistic spectrum sharing or access (OSA) using the
secondary and incumbent channels, and is validated with analysis.
Key-Words: - Cognitive Radio Network (CRNs), Primary users, Secondary Users, discrete-time, Markov chain,
channels.
1 Introduction
The demand for radio band has rapidly been on the
rise. This is not far from the fact that wireless
network are speedily gaining popularity over their
wired counterpart mainly due to the low cost and
portability, which, in turn has exponentially increase
the demand for spectrum [1]. Also, wireless
application and services are rapidly on the rise in
size, number and complexity thereby are band width
hungry which in turn demand for more spectrums.
However, extensive measurement shows that the
fixed frequency allocation results in low utilization
(only 6%) of the license radio band in most of the
time [2] [3]. Moreover the remaining portion of the
unlicensed band is being used-up by these evolving
wireless services and application, hence leading to
the problem of spectrum scarcity and hence call for
a better spectrum management strategies and
policies[4][3].
In other to better maximize the license band, a
promising approach which improves spectrum
utilization, by dynamically allowing spectrum
sharing between the licensed users known as the
primary users PUs and the unlicensed users called
the secondary users SUs is proposed, all in the bid
to cope with rapid growth in multimedia and
wireless services and applications with limited
resources (spectrum) [5] [6]
The Cognitive Radio system uses OSA [7] [5]
which is an optimal way to exploit the temporary
unused spectral resources by constantly sensing the
primary user band. Also the sensing information
which is like a feedback to the cognitive radio base
station (CRBS) of the secondary users form an
awareness map of all the licensed band that is not in
used or idle and so manageability massive unused
resources become less of a problem. But this
primary uses are not aware of the existence of the
secondary users and so, no pre-notice is given
before they arrives to occupy their channels, these
raise the question of how the secondary users will
react when the primary users appears? This paper,
we investigates to show the impact of primary users
on the secondary user performance through
simulations and carefully worked-out model couple
with a simplified discrete–time two state Markov
chain to analyse the PU channels occupancy. The
channel takes the value of 1or 2 subject to whether
the channels are in BUSY-state or IDLE-state
respectively
Recent Researches in Electrical Engineering
ISBN: 978-960-474-392-6
284
2 Related Work
The underutilization of spectrum under the present
fixed spectrum management strategy has spur many
research work especially on dynamics of accessing
the primary user channels [10]. In [7] the authors
proposed a two cognitive MAC protocol scheme to
support voice services in the presence of the primary
user .In [8] the author came up with a simple
approximated model for un-slotted OSA networks
under a non-saturated condition. In [9] the author
classifies secondary users as high and low priority
respectively and buffers the secondary users with
low priority. In [11] the author did a unique
performance analysis but did not consider if it is a
centralized setup and also did not consider delay as
one of his performance metrics. A catalogue of
dynamic spectrum access is found in author [12].
Lastly, the OSA is one of the several approaches to
dynamic spectrum access.
3 System Model and Assumptions
Considered in this work is an infrastructure-based
cognitive radio network scenario where a PU
channel are sensed by secondary user and send
sensed information to the cognitive based station
which grant access to the primary channel see Fig.
1. Note, when the PU is idle or not occupying its
channel the channel becomes that of the secondary
user at that time. Though, when the primary user
arrives, by the principle of overlay, the secondary
user will vacate its channel due to PU high priority
even if SU has packet to transmit.
Fig. 1 Network Model/Architecture
This paper considered a single primary channel and
secondary channel. The time is partitioned in slots,
and each time slot the primary user is either busy
(occupied) or not. The primary channel is at Busy-
state if it is used by primary users, and at idle-state
when is not using or transmitting packets and
secondary users are making use of its channel. The
occupancy of the primary channel is modeled as a
discrete-time two-state Markov chain. The channel
state takes the value 1 or 2 depending on whether
the channel is at busy state or at idle state,
respectively. It transits from Busy-state to Idle-state
with probability and stays in idle-
state with probability . Fig .2a and
2b shows a transition diagram and the sensing of
secondary users. Note that busy/on and idle /off are
interchange respectively.
We considered a secondary user looking for
spectrum opportunities in the primary channel.
However, prior to sensing and gaining access to the
primary user channel, the secondary users initially
harmonize with the slot arrangement of the
incumbent channel. If the primary user is sensed to
be busy, the cognitive users can send its packet. But
if the channel or slot is busy, cognitive users restrain
from sending data to avoid interference. However,
the cognitive users keep sensing the channel/slot by
keeping the packet at the head-of line position of the
queue, and sense the primary channels again at the
next slot see Fig. 2a.The cognitive users senses the
primary channels at an average interval of
sequentially for each time slots at the beginning of
each fixed queue for a secondary users packet for
each slot beginning of the queue, where [11].
However a secondary user have not send packet for
succeeding time slots, it will sent in the next time
slot.
Fig. 2a ON/OFF slots diagram showing SU sensing
Fig. 2b ON/OFF Markov transition diagram
The system assumptions are as follows:
• The secondary channel is dedicated to the
secondary users.
• Precise sensing of primary users channel by
the secondary users.
• The cognitive user operates under overload
condition where it has packet to waiting to be
sent.
• An infrastructure-based cognitive radio
network.
Recent Researches in Electrical Engineering
ISBN: 978-960-474-392-6
285
• The secondary users could be real time users
(packetize voice call like Skype, viber etc.) or
non-real time or elastic (file downloading,
browsing etc.)
• Secondary users transmit in Packets.
4 Performance Analysis of the System
Model
In this section, we analyse the performance of the
system model by using a hidden Markov chain
(HMC). Also, considered, is a fixed interval points
hidden at the start of the time slot after a packet
leaves the queue illustrated in Fig. 2a. Denote the
order of hidden points. Let be
the order of the hidden points and be the
state of the primary channel at the hidden
point .
Then, we define our state as
as a birth death process whose behaviour guided by
the Markov kernel or transition matrix. Expressed
[11] as
For and .the
transition matrix kernel is expressed as:
Where
(5)
Therefore,
is the transition probability form state 1 to 2 or
respectively, for the hidden markov chain
{ expressed as;
(8)
(9)
(10)
(11)
From the problem formulation, we are dealing with
discrete random variable. However, the conditional
probability mass function,
For, given that is gotten as
(13)
In this work, we introduce the service time and the
delay between transitions these two are essential
performance metrics when considering quality of
service of cognitive users. However, the delay is the
difference in time between transitions for example
from ON-OFF with probabilities of (0.2, 0.8, 0.8,
0.2 and 0.5, 0.5) respectively, although, we obtained
the delay from our simulations. The service time
could be express as the time frame required for a
packet to be effectively sent after it is positioned in
the head-of-line [11] of the queue at the inception,
which depends on the state of primary channel when
it move to the head-of line position. Let the state
probability be , the probability of the channel in
state will be where then, the state of
the channel when a secondary user packet moves to
the head-of -line of the queue is expressed as,
Recent Researches in Electrical Engineering
ISBN: 978-960-474-392-6
286
Haven established that the state of the channel is
when the secondary user packet moves to the head-
of-line position of the queue, the conditional
probability that the service time of the packet is is
precisely the same as the probability of .
However, conditioning the state of the primary
channel, the probability distribution can be
determined for the service time of the secondary
user packets. The probability of that the
service time of a secondary user packet is expressed
as,
The average service time of the secondary user
packet is expressed as
Hence, the saturation throughput and which is
the throughput of the secondary user over the
primary channel and the secondary channel
respectively, defined as the number of secondary
users packets successfully sent on the primary
(respectively secondary) channel per slot. Thus, the
saturated throughput and can be expressed as,
The secondary user total throughput of the
secondary user, defined as the number of secondary
user packet successfully sent per slot, is expressed
as,
5 Numerical Result and Discussion
In this section, we present a numerical result to
illustrate the performance of the SU using total
throughput of PU and SU, throughput of SU
Separately, delays, and average service time. All the
illustrated analysis results are based on theoretical
analysis and have been verified to a large extent by
simulations.
Fig.1 illustrate the average service time under
different occupancy statistics of the primary
channel. With a large probability of 0.8, there is
likelihood of the PU state to remain unchanged.
This relates to a long tailed traffic arrivals in the
primary channel [10]. In the second instance, there
are equal chances for it to change state or remain in
its current state and lastly, the state of the primary
users renaming unchanged with a small probability
of 0.2. The time slot λ is varied from 1 to 20. Form
the result in fig.1 it shows that different traffic
statistics of the primary channels have different
performance, but have consistence after a point
along λ that is, as λ increase the average service
time of the SU packet increase and at a point remain
constant. Fig.2 shows that as λ increases, the
throughputs Tp and T increases but the saturation
throughput over secondary channel Ts which is the
difference between Tp and T decreases to 0 as λ
increases as shown in fig. 3 And lastly, in fig. 4 it
shows the delays for each of the probability as it
transits from one state to another. A 0.8 probability
indicate that, there are higher changes for it to spend
more time than other probability but as λ increases
the delay becomes almost uniform indicating
saturation.
Recent Researches in Electrical Engineering
ISBN: 978-960-474-392-6
287
Acknowledgement
This work is supported by Telkom SA and
Alcatel Lucent in the Centre for Radio Access &
Rural Technologies, Centre for Engineering
Postgraduate Studies (CEPS HVDC/smart grid
centre).
6 Conclusion
We have shown the impact of the primary users
on the secondary channel through and analytical and
simulation approach. Also, evaluate the secondary
user through their performance metrics like delay,
throughput and service time. Shown the
performance of an OSA scheme, Work is in
progress is on channel assembling strategies in an
MF-TDMA based CRN, where identified Idle/OFF
mini-slots in a sub-channel are assembled for use by
real-time and non-real time secondary users
respectively.
.
Recent Researches in Electrical Engineering
ISBN: 978-960-474-392-6
288
References:
[1] Bechir Hamdaoui and Kang G. Shin “OS-
MAC: An Efficient MAC Protocol for
Spectrum –Agile Wireless Networks” IEEE
Transactions on Mobile Computing VOL. 7.
NO. 8, AUGUST 2008, Page 915-930.
[2] Linbo Zhai, Kaiming Liu, Yuanan Liu,
Ming Yang, Lin Zhuang. “A Slot-Based
MAC Protocol in Cognitive Radio Wireless
Networks” 4th International Conference on
Wireless Communications, Networking and
Mobile Computing, 2008. WiCOM '08.
2008 , Page(s): 1 - 4
[3] M. Mchenry “Spectrum white space
measurements” New America Foundation
Broadband Forum, June 2003.
[4] Mehrdad Khaledi and Alhussein A.
Abouzeid “Auction-Based Spectrum
Sharing in Cognitive Radio Network with
Heterogeneous Channels.” Information
Theory and Applications Workshop (ITA),
2013 , Page(s): 1- 8
[5] Abdelaai Chaoub, Elhassane Ibn Elhaj,
Jamal El Abbadi. “Multimedia traffic
transmission over TDMA shared Cognitive
Radio networks with Poisson Primary
traffic.” International Conference on
Multimedia Computing and Systems
(ICMCS), 2011 Page(s): 1- 6.
[6] Mitola III, “Cognitive Radio: An Integrated
Agent Architecture for Software Define
Radio” Ph.D. Thesis KTH Royal institute of
Technology, 2004
[7] P. Wang, D. Niyato and H. Jiang, Voice-
service capacity analysis for cognitive radio
networks, IEEE Trans. Vehicular
Technology 59 (2010), no. 4, 1779-1790.
[8] Y. Lee, A simple model for non-saturated
opportunistic spectrum access networks,
IEICE Trans. Commun. E94-B (2011), No.
11, 3125-3127.
[9] Y. Lee, C. G. Park and D. B. Sim, Cognitive
radio spectrum access with prioritized
secondary users, Applied Mathematics &
Information Sciences 6 (2012), no. 5S,
601S-607S.
[10] Q. Zhao, L. Tong, A. Swami and Y.
Chen, Decentralized cognitive MAC for
opportunistic spectrum access in ad hoc
networks: a POMDP framework, IEEE
Journal of Selected Areas on
Communications 25 (2007), No. 3, 589-600.
[11] Yutae Lee “Performance
Analysis of Cognitive Network
with Primary and Secondary Channels”
East Asian Mathematical Journal Vol.
29 (2013), No. 1, pp. 101-107
[12] Q. Zhao and B. Sadler “Dynamic
Spectrum Access: Signal Processing,
Networking and Regulation,” Signal
Processing Magazine, IEEE, May 2007.
Volume:24, Issue: 3 Page(s): 79- 89
Recent Researches in Electrical Engineering
ISBN: 978-960-474-392-6
289
