Enhanced Cognitive Femtocell Approachfor Co-Channel Downlink Interference Avoidance
Sinan Khwandah,John Cosmas, Ian A. Glover, Pavlos Lazaridis,Giuseppe Aranitiand Zaharias Zaharis
Sinan Khwandah and John Cosmas are with the Department of Electronic and Computer Engineering, Collegeof Engineering, Design and Physical Sciences, Brunel University London, Uxbridge, Middlesex, UB8 3PH, United Kingdom (e-mail: {sinan.khwandah, john.cosmas}@brunel.ac.uk).
Ian A. Glover and Pavlos Lazaridis are with the Department of Engineering and Technology, University of Huddersfield, Queensgate,Huddersfield, HD1 3DH, United Kingdom (e-mail: ).
Giuseppe Aranitiis with the Department of Information Engineering, Infrastructure and Sustainable Energy (DIIES), University Mediterranea of Reggio Calabria, LocalitàFeo di Vito, Reggio Calabria 89100, Italy (e-mail: ).
Zaharias Zaharis is with the Department of Electrical and Computer Engineering, Aristotle University ofThessaloniki, 54124 Thessaloniki, Greece (e-mail: ).
Index Terms—LTE, Femtocell, Cognitive Radio, Co-Channel, Interference, CQI, MCS.
Abstract
The deployment of low-power nodes such as femtocells within the macrocell’s coverage area is one of the main features of future 5G networks. Although these small nodes increase the system capacity and improve the link quality, there will be challenges in the femtocells planning and management particularly due to the occurance of interference when they are deployed in a co-channel with the macrocell. The cognitive radio inspired femtocell is one of the solutions to mitigate interference.In the presented paper,the femtocell is made self-organised and performs uplink sensingto enhance the cognitive information. In more detail, the femtocelldiscovers the interference-free resources of far macro-users; in addition, it exploits the channel dependent scheduling to predict the future resources of the nearby victim macro-users andavoidsutilising them.Simulation results show that the co-channel interference on the macrocell Downlink (DL) signalis minimal, and therefore performance improved, under the condtion that cognitive femtocelloperates adaptively and monitors the spectrum allocations countinuously.
I.Introduction
The mobile phone industry has undergone significant evolution and introduced innovative services, leading to a competitive market. Operators are always looking for new technologies which help to serve this increasing growth. One of the current methods is to use broadband internet to increase radio coverage by deploying the femtocell technology.Femtocells are beneficial to network operators and users alike. For the former, offloading traffic from macrocells down to smaller base stations increases network capacity [1]whereas for the users, femtocellsimprove the User Equipment’s (UE) reception and the network availability. As studies have shown that the largest number of phone calls are still made from indoors [2], femtocells can potentially bring large benefits to indoor wireless communications. One femtocell provides coverage and support to a small number of users by utilising the customer’s internet backhaul such as Digital Subscriber Line(DSL) to connect to the cellular network.Due to their low power transmission, femtocells reduce the amount of interference with other electrical devices, whilst providing improved indoor coverage. In addition, the battery life of the mobile device can be prolonged as it only needs to connect over short distances when connecting to a femtocell, instead of to the nearby macrocell [3].
However the presence of small base stations within the coverage of larger ones changes the architecture of the cellular system. Network operators may prefer co-channel deployment as it provides a higher system capacity, but it is also most likely to lead to interference[4].
Femtocell technology is evolving, and the effects of femtocells on non-femto-users such as macro-users have been a subject of much research.Different approacheshave been proposed in order to reduce and mitigate the co-channel interference. The Almost Blank Subframe (ABS) is one method to enhance intercell interference coordination, which is presented in [5],where the DL scenario was analysed in order to find the required number of ABSs based on modelling the base station and UE placements. Although the ABS method manages the interference well, it causes time-frequency resource waste for the macrocell.Power control is exploited widely in 4G systems in order to reduce the effects of dead zones created by the femtocell coverage.One approach for power control such as thatpresentedin [6],where a distributed femtocell DL power control is based on the minimum power required for the femto-users to meet the DL outage probability.However, the major drawbackin power control schemes is the degradation of the SINR at the femto-user’s sideif the femtocell reduces its transmitted power by a large amount in order to reduce the interference level.
The cognitive femtocell has been proposed in many works such as in [7],where the femtocell is trained to predict the traffic pattern of the macro-users in order allocate the spectrum opportunities. Another cognitive approach,such as the one presented in [8],investigates opportunistic co-operation between the macro-users and the femto-user so that the femto-user cannot transmit concurrently with the macro-user and it should follow pre-defined models based on macro-users’ idle and busy periods.
Although cognitive femtocell is a promising technology, employing the cognitive approach to find the access opportunities is not a trivial task. In LTE systems there is flexibility in assigning resources to the usersandallocations vary in response to the channel condition at the user location.In heavily loaded big cells such as the macrocell,scheduling is highly dynamic and changes every subframe so the allocated resource blocks in one subframe maynot be allocated again to the same user in the following subframe.This causes instability at the cognitive femto-scheduler and degrades the QoS as the cognitive femtocell cannot sense and utilise/evacuate the resources simultaneouslyat every millisecond (subframe). Thus thecurrent cognitiveinformation of the occupied macrocell resources may not be usefulenough unless it is enhanced by knowing the future candidate resources.
Therefore, in this paper we propose a method for robust sensing performed by the co-channel femtocell in order to be aware of the current and future scheduling opportunities. The current cognitive approach comprisesof the DL listeningto the macrocell signal to find the current scheduling and free resources, and UL sensing to discover the remote macro-users resources such as the approach studied in[4]. The superiority of the presented scheme is that the femtocell in addition to listening to the macrocell DL signal and sensing the UL signal to discover the far macro-users, it is enhanced to decodethe nearby macro-user ULsignal topredictfuture scheduling opportunities depending on the link adaptation processand channel dependent scheduling in LTE. There is minor co-operation with the macrocell over the X2 interface, which will be explained in the following sections, in addition to demonstrating the advantages and possible drawbacks of the presented work. Where is the novelity
The remainder of this paper is organized as follows: in section two the proposed scheme is presented and the importance of interference management is demonstrated in addition to the co-channel deployment challenges and macro-user tracking. In the next section, interference avoidance performance is evaluated by presenting the performance results of conducted simulations.
II.Proposed Scheme
a.Requirements
It is vital to the efficiency of the system that the performance of femtocell does not undermine the activity level of the primary users, namely: the macro-users. A feature of these low power nodes is that they are deployed without planning furthermorethey can be deployed by the user at any time. Therefore, any interference management scheme must afford priority to the macrocell and its users.
The random installation by the end-user and the ad-hoc nature of the femtocell positioning indoors makes it difficult for the cellular operator to control the planning of these new low power base stations (the HeNB) without any kind of self-control or intelligence added to them. The changing environment and also the location of the femtocell, make it very important to equip the femtocell with some kind of sensing capabilities. Therefore, it would be more efficient if femtocells are planned and introduced into the market with self-configuration and self-optimization capabilities.
b.The co-channel deployment
Each LTE subframe consists of a control region and a data region. Co-channel interference happens when there is collision between the control region of the transmittedsubframes of the macrocell and the control regions of the other transmitted frames from a femtocell operating on the same frequency channel.The control signals are randomly distributed within the control region and they represent the scheduled UEs. If the control region is corrupted, the UE will not be able to decode the information contained in that region to find its resources andas a result it will not be able to access the network and thus it is considered as a blind. Moreover, in the case of severe interference, the UE may not be able to connect to the serving eNB (femtocell or macrocell).
Fig.1Femtocell and Macrocell frame timing.
If the data region is corrupted in some places, the Hybrid Automatic Repeat Request (HARQ) mechanism could help in restoring the corrupted data through multiple re-transmissions.Block Error Rate(BLER)of 10% is the critical recommended value for a successful reception of the BLOCK and shouldnot be exceeded [9]. However, there is no HARQ for the control channels, which span the entire bandwidth, therefore they must be protected and the target BLER over them must typically be 1% or less.
In order to avoid interference or reduce it to the lowest possible value over the control channels, the femtocell frame is shifted by threesymbols relative to the macrocell’s DL frame.Fig.1 shows the frame timing of the femtocell and the macrocell, assuming that the femtocell and the macrocell are time synchronised. This is similar to what is presented in 3GPP technical report in [10]. Another measurethat could be performed by the femtocell is reducing the control region size from three symbols to one when the number of connected femto-users issmall, but this procedure applies only when the femto-users have low activity level and require low number of resources.
Regarding interference mitigation over the data region, in theabove-mentioned3GPP report [10]they rely on link adaptation process and channel dependent scheduling in LTE to mitigate the overlap over the data channel where the victim macro-user can request scheduling over resources on which it receives strong signals and are not occupied by the femtocells. However the efficiency of the channel dependent scheduling depends on the availability of the vacant resources as will be illustrated in the following sections.
c.Sensing
Thefemtocell self-configuration procedures mainly include spectrum sensing. The radio frequency measurements performed by the femtocell are considered similar to the measurements performed by the UE in the surrounding area of the femtocell. Many factors may affect the accuracy of the performed measurements such as the power of the sensed signal, switching the femtocell off, and changing its location from a place deep inside a building or close to a window. Therefore, the femtocell has to keep performing the radio measurements in order to be aware of its surrounding radio environment.
For spectrum sensing and detection, many options could be utilized such as defining a power threshold to determine theUplink (UL) signals of interest. The threshold value depends on how successful the femtocell is at finding resources. A higher threshold means that the femtocell selects the macro-users which are very close, whereas a lower threshold implies that the femtocell considers more macro-usersto avoid their allocated subcarriers, which results in a low number of subcarriers being left to be utilized by the femtocell andassigned to theirfemto-users as those subcarriers are assigned to the nearby macro-users and could not be used by the femtocell for the interference avoidance considerations.Then, the femtocell identifies the close macro-usersto findtheir DL resources and avoidsutilizing them. The femtocell may discover free subcarriers (unused by the macrocell), in addition to subcarriers allocated to distant macro-users. Femto-users are limited in number compared to macro-users and this means that there are alow number of subcarriers that need tobe allocated tofemto-users. By judiciouslyselecting the subcarriers, the femtocell interference on the macrocell DLcould be avoided.
d.Search space and macro-user tracking
Fig.2a describes the proposed algorithm for interference avoidance and macro-user tracking. After start up, the femtocell performs listening using a listening module to receive the DL reference symbols of the macrocell. This is not straightforward and requires the femtocellto be synchronizedwith the macrocell [11].Once the synchronization signals from an eNB are audible, this eNB needs to be evaluated. The femtocell detects the frame timing from Primary Synchronisation Signal (PSS) and Secondary Synchronisation Signal (SSS) signals, and the DL bandwidth from the Master Information Block (MIB). Once the femtocell discovers the macrocell bandwidth and the starting frequency fM (Fig.2c), it can discover if there is any resource overlap with the macrocell.
The percentage of overlap depends on the bandwidth allocated to the femtocell. The femtocell mayshare the whole macrocell’s spectrum(full spectral usage) or the femtocell is allocated smaller bandwidth (partial channel sharing) (Fig.2b). After synchronization and system discovery,if the femtocell discovers that there is an overlap with the macro-resources, it has to sense the macro-users presencewithin its detection areaand perform interference avoidance. The nearby macro-users could be determined based on the Received Signal Strength (RSS) [12], and their uplink signals are captured and demodulated through removing the cyclic prefixes and performing FFT on the received subframes.
The UE can be identified in the LTE system by different keys, whilstit is connected with its serving eNB. UE identifiers are allocated to the user duringRadio Resource Control RRC_Connection setup [13]. A unique identifier, the Cell Radio Network Temporary Identifier (C_RNTI) is allocated by the eNB to the UE at the moment when an attach procedure is initiated between the UE and the eNB. The UE depends on the C_RNTI to track its data on the DL within the cell and this makes the C_RNTI a very important identifier for marking RRC messages during connection. Another identifier such as the Radio Network Temporary Identifier (RNTI) identifies the Physical Downlink Control Channel (PDCCH) allocations, which carry the Downlink Control Information (DCIs). This identifier is used by the UE when it monitors the search spaces in order to distinguish between a common search space and an UE specific search space. The DCIs with general network information are broadcast on a common search space such as paging, uplink power control and random access, while the DCIs with user-specific allocations are carried in the UE’s specific search space.
Fig.2Interference avoidance scheme
The C_RNTI thus defines uniquely which data is being sent in a DL direction within a particular LTE cell that belongs to a particular UE. The nearby macro-user discovery is performed on the basis of a C_RNTI. Thereby itis made possible for the femtocell to discover a relatively considerable number of macro-users in its vicinity.
e.Future resource prediction
The femtocell performs two tasksupon sensing the UL signal of the macro-user of interest; first it locates the macro-user’s C_RNTI in order to find its resources on the macrocell DL map. This gives the femtocell a preliminary vision of the macrocellDL free resources in addition to resources allocated to far macro-users. Secondly, the femtocell discovers the macro-user’s preferred subband for the next scheduling.
The femtocell performs calculations to locate the interference-free resources; i.e. free resources and resources allocated to the far macro-users. Calculations are based on the smallest allocation unit in the LTE system, which is the resource block (RB) (12 subcarriers, each subcarrier is 15KHz). The bandwidth is divided into groups of subbands and each subband includes a number of resource blocks according to the bandwidth size as shown in Table-I.
Table-I LTE subbands and their resource blocks
BW (MHz) / Num. ofRBs / Num. of Subbands / RBs/
Subband / RBs/Last Subband
1.4 / 6 / 1 / 6 / -
3 / 15 / 4 / 4 / 3
5 / 25 / 6 / 4 / 5
10 / 50 / 8 / 6 / 8
15 / 75 / 10 / 8 / 3
20 / 100 / 13 / 8 / 4
The femtocell uses the decoded UL information to predict the candidate subbands for the nearby macro-users by exploiting the channel dependent scheduling process. The macrocellwill try to schedule the macro-user on its preferred subband through the channel dependent scheduling. Based upon the SINR measurements on the DL, the macro-user reports a Channel Quality Indicator(CQI) value for each subband including the wideband [14, 15]-to be translated to an appropriate Modulation and Coding Scheme(MCS) index- based on the target link quality in terms of DL BLER.The wideband’s reported MSC value is used by the serving macrocell to optimize the frame resourcesand differentiates the best reported subband by its UEs, by calculating the metrics for the subbands, which depend on the difference between the reported subband MCS and the wideband MCS, and the allocation blocksize.The UEmay span multiple subbands, and some of those subbands can be quite bad. If anysubband goes below the wideband MCS index,it is taken out of consideration for savings.The femtocell finds which subband the CQI corresponds to, it obtains this information from the initial Physical Uplink Control Channel (PUCCH) offset and CQI transmission parameters such as the periodic configuration index (which controls the periodicity of the CQI reporting over PUCCH -the number of subframes between 2 consecutive CQI reports) and the subband report repetition count (This parameter controls the periodicity of subband reporting before the next wideband CQI is reported)[16].