Cognitive Radio Based Femtocell Reource Allocation Computer Science Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Femtocell network provides the solution to the dead zone problem which occurs in the indoor environment. Heterogeneous interference occurs as a result of deployment of femtocells in already existing macrocells. Heterogeneous interference can be minimized by dedicated channel deployment where femtocell makes use of different frequency assignments from that of the macrocell, but this is impractical in thick femtocell distribution. Therefore co-channel deployment is used and it is possible by adaptively varying the transmission power. Inter-cell interference in a macrocell can be reduced or avoided by:

Using fractional frequency reuse where the frequency band is divided into frequency assignments which contains sub-channels and adjacent cells are allocated different frequency assignments. Heterogeneous interference can be reduced in this case by assigning different frequency assignments to femtocells from the macrocells they exist in. since femtocells may be massively positioned within the macrocell; heterogeneous interference will have to be reduced by adjusting transmission power.

Using coordinated beam forming and joint antenna processing techniques.

Cognitive radio technique was applied for resource allocation by the femtocell to exploit the interference free channels by cognitively recognising the downlink interference signature but its disadvantage is seen when the power detected in a channel by the femtocell is very low and it mistakes the channel for an interference free channel.

The femtocell independently manages heterogeneous interference by determining an occupied channel through estimating its uplink energy and assigning unoccupied channels to its femto-users. When a femtocell detects a macro-user in a sub-channel it vacates the entire frequency assignment. The uplink energy is estimated instead of the downlink energy because the uplinks transmit power is usually larger than that of the downlink. Also the femtocell selects the frequency assignment with the lowest signal energy if no vacant frequency assignment {that is at least one of its sub channels is in use} is available.

When allocating sub-channels the femtocell base station estimates the downlink interference signature of the operating femtocell user for all the sub-channels in its frequency assignment and then sorts it in increasing order the first sub-channel on the list is then assigned. The femtocell base station adjusts the transmission power of the allocated sub-channel to mitigate interference to adjacent femtocells. Applying cognitive radio technique to femtocells maximizes throughput and minimizes interference this is done via channel sensing and resource scheduling.


In managing cellular spectrum, the possible methods which could be used include independent interference control, network self-management and intelligent network configuration. The random deployment of femtocells into an already existing macro cellular area requires interference management to be performed in an independent and flexible way according to specific infrastructure topology and traffic distribution.

Cellular intelligence is required to optimize network management and this is implemented through smart power control and adaptive resource allocation methods. The speed of power control is determined by the channels coherence time of the communication link, in femtocells coherence time is usually longer because the channel does not vary quickly.

For interference management femtocells are considered as secondary systems and orthogonal channels are independently allocated to prevent interference to macrocell communications. To implement this spectrum reuse procedure is employed and the level of reuse is determined by the frequency reuse factor. The frequency reuse factor is determined by how much co-channel interference the receivers in the network can bear.

Cognitive femtocell is when the channel reuse factor is cognitively determined depending on the femtocell channel environment and this is a flexible reuse scheme. Interference in femtocell is managed in a cognitive way by introducing interference recognition and environment perception. The femtocell senses the interference level across all cellular channels and indicates it with its interference signature and the channel which has minimum interference to the network environment is allocated. For optimal channel reuse pattern the channels are divided into three types:

Interference free channels: these channels experience zero interference to them so the frequency bands in this category have a higher reuse priority.

Soft interference channels: they have interference caused by far away femtocells and the best channels with small co-channel interference can be used.

Hard Interference channels: the frequencies in this channel cannot be reused as the experience strong interference from neighbouring cells. The femtocell cognitively reuses channels orthogonal to the hard interference channels.

To implement independent interference management the cognitive radio concept is used where software defined radio terminal senses the available frequency channel cognitively and it then accesses the spectrum opportunistically.


Cognitive radio enables dynamic spectrum access thereby resolving the problem of spectrum scarcity and underutilization. They are capable of recognizing and collecting useful information about the frequency channel and also adapt their transmission schemes to achieve the best performance.

Two basic operational models for cognitive radios are:

Opportunistic spectrum access: the secondary users in this case utilize the frequency channel for transmission when it's not being used by the primary users this is made possible by spectrum sensing. This method is primary-transmitter conscious.

Spectrum sharing: the primary and secondary users in this case transmit within the same frequency channel simultaneously with the condition that the secondary users control the consequent interference at the primary user's receiver. This is primary-receiver conscious. It utilizes the spectrum more efficiently by allowing orthogonal transmissions by the primary and secondary users. It is also more effective for managing interference.

In the design of spectrum sensing cognitive radio the decentralised approach is used where primary users network is designed without knowledge of the secondary user's network and the secondary user's network is designed with limited knowledge of the primary user's network.

There are two standards for decentralised approach spectrum sharing based cognitive radios:

The secondary user employs part of its power to transmit its message and the remaining power to relay the primary user's message therefore compensating for the interference it causes at the primary user's receiver. This is known as cognitive relay concept.

The secondary user is imposed a maximum secondary user interference power constraint at the primary user's receiver. This constraint is the IT constraint.

Dynamic resource allocation technique for cognitive radio helps for optimal deployment of transmit strategies to maximize the secondary network's throughput and based on channel state information of the primary and secondary user's network; the bandwidth, transmission power, bit-rate and antenna beams are dynamically allocated.


Femtocells are home base stations which are short range (small cellular sizes), low cost, and low power base stations which provide better indoor voice and data coverage. It was seen that a 1600x gain was achieved through reduced cellular sizes and this improved gain was obtained from efficient spatial spectral reuse or high area spectral efficiency. This is the major benefit of femtocell. The reasons for the development of femtocell networks are:

It provides reduced direct costs for the service provider.

Most voice calls (50%) and data traffic (70%) originates from indoors.

The advantages of femtocells include:

Femtocells helps to provide improved reception and high capacity due to its short transmit-receive distance (also minimized path loss exponent) which then leads to increased battery life, low transmission power and high SINR.

They also provide improved macrocell reliability since the indoor user traffic will be implemented in the femtocell therefore the macrocell mobile users will experience better reception due to reduced traffic.

Also femtocell usage reduces the operating and capital costs for the operators.

The benefits in spatial reuse are obtained in exploiting diversity and making use of interference management techniques (that is, suppression, cancelation and avoidance). Major interference challenges in femtocell networks will occur due to macrocell - femtocell interference, femtocell - macrocell interference and femtocell - femtocell interference. Interference avoidance is the best way of interference management because compared to the other techniques; this does not have the probability of loosing part of its information in the process of managing the interference. Interference avoidance is best provided via the use of directional antennas, frequency & time hopping and adaptive power control where power is adjusted to avoid cross tier interference. For example; reducing transmission power as distance to macrocell base station reduces.

The problems facing broadband femtocells are:

Resource allocation: the already existing macrocell base stations are placed by the operators and therefore they take into account the centralized frequency planning to avoid interference while in femtocells the consumers place the base stations without knowledge of frequency planning to avoid interference.

IP backhaul: this must provide quality of service to delay sensitive traffic and sufficient traffic capacity to avoid bottle neck. Femtocells should also be able to handle WI-FI traffic, data and voice services simultaneously.

Timing and synchronization: femtocells require timing and synchronization so as to align the received signal and thereby minimize interference and also to ensure tolerable carrier offset. It is also required so that effective handoff could occur between femtocells and macrocells. To achieve timing and synchronization in femtocells the following solutions may be implemented:

Using high precision oven controlled oscillators.

GPS for synchronizing femtocells with macrocells.

Network solutions with timing accuracy of 100ns and self adaptive timing recovery protocols.

T he challenges encountered by voice femtocells include:

Allocating frequency bands to macrocells and femtocells. It could either be allocated with the thought of reducing interference by assigning different frequency bands to the different cells or to with the aim of spectral efficiency by allocating same frequency bands to cells within the environment.

Providing open access or closed access to the femtocells networks: open access implies that users within radio range (better coverage) of a femtocell will be allowed to use it while closed access implies that the femtocell network is only used by those licensed to use it (private and secure).

MIMO femtocells enable the femtocell network implement link adaptation transiently thereby allowing femtocells to switch between robust transmission and high data rates. High data rate is achieved by utilizing spatial multiplexing over high SINR links while robust transmission by using open and closed loop diversity schemes over low SINR links.


Green communication points to ways of power efficient communication and making use of the already available excess bandwidth and also finding methods to reduce device complexity. Cognitive radio for green communications implies using the free spectrum present with power efficient modulation. The benefits of minimizing energy usage are:

Environmental friendly - green communications

Allows the use of smaller batteries in devices - battery life lasts longer

Allows transmission of higher amounts of data because capacity does not increase linearly with power.

A major part of cognitive radio is spectrum sharing and the ability to adapt by shifting between different modulation schemes. Power efficiency can be increased that is power usage reduced while keeping data rate or capacity the same if the available bandwidth is fully used as seen in the Shannon capacity equation:. It is seen that capacity increases linearly with bandwidth but logarithmically with power. Without the use of cognitive techniques, improved bandwidth efficiency is seen to reduce battery life and vice versa but with cognitive techniques battery life is not affected by bandwidth efficiency.

Power path gain ratio measures the effect of a new transmitter on the present receivers; this ratio serves as a threshold for moving between spectrum splitting and spectrum sharing to maximize power efficiency while minimizing the effect on bandwidth efficiency. The power path gain ratio is given as:.

A cognitive engine with power efficient schemes can be more bandwidth efficient when target data rate increases. The cognitive engine is equipped with a learning algorithm and this algorithm helps to find the optimal power path gain ratio threshold so as to decide on which spectrum strategy to use (splitting or sharing). The chosen spectrum strategy is then stored in the knowledge base. The reasoning engine uses the information in the knowledge base to determine the optimal modulation scheme to be used and also the power/ bandwidth efficient strategy according to the spectral occupancy probability. The occupancy probability is used in power efficient systems so as to make sure that the system is bandwidth efficient when the spectrum is congested.

Bandwidth efficient systems use the highest level modulation scheme when the percentage of unoccupied channel is minimum to obtain the target data rate while power efficient system use the low level modulation scheme when we have a high percentage of unoccupied channels. Power efficient systems have increased bandwidth efficiency when occupancy probability is above 60%.

The reasoning engine of the power efficient system shifts from low modulation scheme to high modulation scheme as occupancy probability and SINR increases. Power efficiency reduces in a power efficient system as a higher modulation scheme is used when the maximum number of channel increases so as to achieve target data rate. The reasoning engine of the bandwidth efficient system uses the highest modulation scheme possible to obtain bandwidth efficiency.

Methods of improving energy efficiency include:

Through antenna directionality: This reduces antenna transmission power and interference by keeping data rate and range of communication fixed.

Using distributed artificial intelligence to enable a more efficient method of spectrum sensing

Using limited intelligence with low complexity sensing techniques to provide good performance.

6. Multiuser WiMAX Resource Allocation Algorithm Based On Cognitive Radio.

Cognitive radio is used to solve the problem of spectrum scarcity in WIMAX deployment. It is an intelligent radio that detects its environment instinctively and then modifies its parameters to suit the environments characteristics. Spectrum resources are allocated by sensing its availability and the level of interference. For power to be allocated to the secondary users the interference these users cause to the primary/ licensed users is considered and the power allocated must not exceed a certain level. The interference threshold is set as a function of the signal to interference plus noise ratio of the primary user.

The main aim of resource allocation is to use an allocation technique in which the frequency resources available are allocated maximally while using minimum power but with the quality of service condition still met. OFDM resource allocation allows the secondary users make use of the available spectrum without causing interference to the primary users. Water filling algorithm is a good allocation algorithm to minimize transmission power. Channel state information is needed as this helps in the power allocation algorithm. In the water filling algorithm the sub channels to used by the different users are listed and also the capacity of the channel is known, with this the transmit power for each user is determined. In multi-user allocation, a greedy power allocation scheme is used where additional information of the user's data rate is used in the allocation.

7. Cooperation Based Resource Allocation for Improving Inter-cell Fairness in Femtocell System.

OFDM is chosen as the multiple access scheme because it provides efficient service for multiple users. In OFDM users encounter various channel states so adaptive sub channel allocation can be used to obtain multiuser diversity and also higher spectral efficiency. To maximize capacity sub channels should be allocated based on sub channel gain and power allocated using the water filling algorithm

In iterative water filling algorithm power allocation is performed repeatedly by a user to optimize its own capacity without consciousness of the interference it causes to other users around it. It is a distributed algorithm and its power allocation is carried out without taking the interference it causes to other users into account.

In modified iterative water filling algorithm the user considers the interference to other users and takes appropriate steps to limit this. As a result of the steps taken overall capacity is optimized. It obtains a good performance in interference limited environments but it those not guarantee inter-cell fairness which is a disadvantage. It also does not meet the quality of service requirement for heavy traffic cell.

The enhanced modified iterative water filling algorithm guarantees high overall capacity in high traffic cells while cells with heavy traffic are guaranteed inter-cell fairness. In the EMIWF, sub channel allocation is carried out and using information gathered from it power allocation is performed. Sub channel allocation is performed using proportional fair scheduling and this helps to achieve high overall capacity and fairness requirement guarantee. The power allocation is performed using the lagrangian optimization problem technique while considering the power limitation. The EMIWF algorithm is performed repeatedly to discover channel state variations.

8. Algorithms for Optimal Resource Allocation in Heterogeneous Cognitive Radio Network.

Two techniques for the implementation of cognitive radio where the secondary users are allowed to access the licensed frequency bands without causing interference to the primary users are:

Spectrum Overlay: the temporal available spectrum is used without causing interference to the primary users. It could also be known as opportunistic spectrum access.

Spectrum underlay: secondary users are allowed to access spectrum of the licensed primary users but with a limitation on their transmission power. It is also known as spectrum sharing.

This paper focused on spectrum underlay. The secondary users transmit when the primary user is in the uplink mode but when stop transmission when the primary user is in down link mode. The reason for this is so that the secondary users can collect information on the channel state between them and the base station and this information is necessary for adequate interference management.

A central system in the secondary network collects information from the secondary users and based on this information resources are allocated. The resources are allocated while considering the quality of service requirement. The allocation problems are set in order of priority and based on this priority, resources are allocated; the problem with least importance suffers the highest cost.

9. Resource Allocation with Load Balancing for Cognitive Radio Networks

Cognitive radio allows for accessing licensed spectrum opportunistically without causing interference to the primary user so as to solve the spectrum shortage problem. Load balancing helps to spread the cognitive radios load over the frequency spectrum. This helps prevent overcrowding of the spectrum while another part is being neglected, it also helps with the overall capacity.

Secondary users send the information about their maximum transmit power and desired data rate to the base station and this helps the base station with resource allocation. To perform load balancing, load factor information of the entire frequency band is required and depending on this information users are allocated to the bands. There are different ways of performing load balancing:

Users are assigned to the frequency bands depending on the bands load factor. The band with the lowest load factor is assigned users until it reaches a certain acceptable level, then the next band is loaded until all the bands or users to be assigned are finished.

Users request for frequency band is sorted in descending order depending on their data rate while load factor is in ascending order. The users are assigned depending on their data rate and also on load factor.

This paper recommends a constant power water filling algorithm in the user with the maximum signal to noise ratio is assigned until the channel requirement of that user is satisfied. This is the advantage of multi user diversity because the user with good signal characteristics is assigned channels and users with poor characteristics are given low priorities until they improve. This allocation has the disadvantage of fairness, because users with constant poor characteristics may not be allocated channels.

10. Cooperative Resource Allocation for Guaranteeing Inter-cell Fairness in Femtocell Networks.

Spectral efficiency is required to achieve high data rate. Power control in OFDM helps to reduce inter-cell interference and also maximize spectral efficiency. Inter-cell interference in Femtocell is as a result of its dynamic deployment which then leads to overlapping.

Modified iterative water filling algorithm could be used to solve the power control challenge in multiple cells. Power is allocated to a user in such a way that interference to it is reduced. Interference to a macrocell and Femtocell base station is affected by path loss and this depends on the distance between them.

Enhanced modified iterative water filling algorithm allocates sub channels according to a proportional fair schedule which maximizes capacity and then power is allocated depending on the number of users. The Femtocell base station assigns power such that interference to a nearby Femtocell which has low capacity or heavy traffic is reduced. Increased proximity (reduced distance) between the femtocell base station and the femtocell users reduced the interference contribution because due to the proximity their transmission power is lower.

11. Cognitive Uplink Interference Management in 4G Cellular Femtocells.

Femtocells use adaptive channel reuse depending on its location in the macrocell. The femtocell base station identifies the interference signature for the network and it cognitively uses a channel pattern to avoid interference to and from the macrocell users. The methods for channel reuse are:

Cognitive Channel reuse based on orthogonal scheduler: For the 4th generation cellular network OFDMA and MIMO has been proposed. Using cognitive method in the OFDMA network allows orthogonal channels to be reused by the femtocell without causing interference to the macrocell users. Therefore a femtocell can identify the networks interference and then make use of interference free channels. This gives a reduced channel reuse efficiency because some of the channels in the femtocell will be not be used because of the level of interference the macrocell causes to the channel. A power threshold is set and this defines the channels which cannot be reused. It also defines the femtocell size at which it starts to share channel resource with the macrocell due to orthogonal channelization. The radio sensitivity of the femtocell base station is also defined by the power threshold and the sensitivity factor shows how the femtocell responds to its interference environment. The femtocell determines the channel reuse pattern which produces the best interference signature that is the least interference.

Cognitive channel reuse based on opportunistic scheduler: the transmission power is independently configured to minimize interference to the network. The femtocell base station senses the uplink environment of the macrocell determines its interference signature of the macrocell users and uses this information in opportunistic channel selection. The femtocell base station determines the best channel for reuse that is the channel with the lowest interference. For the power to be allocated, the quality of service requirement of the user is assessed.

12. Cognitive radio resource management for QOS guarantee in autonomous femtocell networks.

The challenges of Femtocells include:

Effective interference reduction

Optimal radio resource usage

QOS guarantee in terms of statistical delay

To reduce interference, dedicated channel deployment could be used where the macrocell and femtocell networks use different channel assignments but this is impractical in dense femtocell deployment because it would result in the femtocell users being allocated very little bandwidth. Therefore co-channel deployment is used where the macrocell and femtocell share all the available spectrum resource.

Interference is reduced if WCDMA is used as the dynamic power adaptive scheme but WCDMA is not favoured by the new technologies such as 3GPP LTE and WiMAx which recommends OFDMA.

Cognitive radio technology requires that the femtocell performs periodic channel sensing to estimate the channels usage by the macrocell. To provide the QOS guarantee the effective capacity theory is used. This is a link layer capability which indicates the maximum constant arrival rate that the system can support so as to ensure the QOS requirement.

The different methods of cognitive resource management in femtocells are:

The femtocell base station senses the channel to determine occupied frequency bands by the macrocell. During the sensing period it senses the received interference power and if its above a certain level the channel is categorized as occupied else it is unoccupied. The femtocell base station then allocates the unoccupied channels to its femtocell users.

The femtocell base station calculates the effective bandwidth which indicates the maximum constant service rate needed by an arrival process to deliver the QOS requirement for real time traffic of a femtocell user.

Using the random scheme for the allocation of frequency channels to the femtocell users, random channels are allocated but the probability of interference occurring is higher which also affects overall capacity negatively. Cognitive resource allocation could be seen as a form of resource waste due to the sensing required but this controls interference which has a positive outcome on the overall capacity. This capacity advantage is due to the ability of the cognitive scheme to recognise potential interference channels.

The cognitive resource allocation technique provides the following solutions:

Enables independent interference reduction even in dense femtocell deployment

Provides statistical delay guarantee

Optimizes radio resource use.

13. Cognitive Femtocell Networks: An Overlay Architecture For Localized Dynamic Spectrum Access

Dynamic spectrum access facilitates the cognitive radio device to analyze the frequency bands and access them if unoccupied until the arrival of the primary user. The primary user refers to the user with greater importance or legal rights to utilize the spectrum band. Femtocells improve spatial coverage with increasing spectrum reuse. They also have the advantage of low complexity and cost. They have increased transmitter-to-receiver proximity which improves system capacity through high quality links and gives more spatial reuse.

Due to the lack of capacity and performance planning in Femtocells which is done in macrocell networks, interference could occur between neighbouring Femtocells and macrocell thereby leading to capacity decrement and performance decline. Advanced power adaptation and synchronization signalling between femtocells are interference reduction techniques.

Opportunistically allocating resources to a femtocell using cognitive technique leads to optimal spectrum reuse. The cognitive based femtocell has spectrum sensing, interference management and resource allocation capabilities. A radio environment map is designed and the information from this helps to control the interference to neighbouring femtocells and the underlying macrocell.

Cognitive femtocell provides higher throughput, adaptive operation, better interference management, lower energy consumption and more reliable sensing. Traffic prediction and modelling is necessary for efficient spectrum allocation. Cognitive radio nodes can estimate traffic using:

Alternating renewal processes where the up and down periods in a renewal cycle which corresponds to busy and idle channel respectively are monitored.

Estimation of call arrival rate and call holding time using an algorithm which makes use of the periodicity of the primary user's traffic.

Depending on this primary traffic estimation channels could be allocated.