Comparative Analysis Of Handover 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.

In this paper we present the comparative abstract of the handover technology. There are always increasing demands of high-speed data applications in wireless systems with fast and seamless access to voice and multimedia services and QoS guaranteed such as in Fourth Generation (4G) Long Term Evolution (LTE)- Advanced system. We would discuss the performance, handover techniques, requirements and features, handover types (soft and hard), handover enhancement in LTE advanced relay networks, architecture and at the end the performance evaluation of transmission control protocol (tcp) and user datagram protocol (udp) during lte handover.

Keywords Lte handover, advanced handover techniques, 4G-relay network, combined handover, femtocell, macrocell, Qos.

Introduction 1. Mobile high-bandwidth data communication is challenge for the telecommunication service providers in this era. User want to access network services in anytime at any place with high speed from their mobile devices. IEEE 802.11 wireless technology is developed to fit user's mobility requirements. With comparison to IEEE the newest technology introduced by telecom vendors by name of LTE (long term Evolution) which is competing access network technologies in 4G wireless network with WIMAX. It is backward compatible with GSM/UMTS cellular systems, this make LTE deployment much easier then WIMAX. [1][6]

2. There are many challenges have to face for enhancing the mobility system in LTE advanced network. There would be two main handover technologies in wireless communication systems hard handover and soft handover. In LTE advanced system carrier aggregation technique is also supported, that eNB could configure a set of cells for UE currently under its service extend. Therefore it always consists of one primary cell and one or more secondary cells. [2][7]

3. For the strong demand of multimedia services and broadband applications with higher data rate the IMT advanced system has been initiated for the next generation mobile communication system 4G.LTE advanced system supports high mobility speed upto 500km/h moreover lte advanced expected to support high data rate upto 1gbps in downlink and 500mbps in uplink. The handover in lte-advanced system is purely based on hard handover. With the limitation of the existing handover technique in LTE system, it is necessary to determine a new handover technique to achieve these goals which include avoiding loss of data, ensure efficiency, reduced outage probability and to increased reliability of handover procedure for the future next generation LTE -Advanced system. [3][8]

4. LTE is designed to provide higher throughput and lower latency than its predecessor, i.e. HSDPA (high speed downlink packet access) which is also known as 3GPP release 5 [3]. LTE also introduces the all-IP network and all services are provided by IP. Therefore the throughput and delay performance must satisfy certain criteria even in the mobility scenario.In addition, there is a link interruption in the user plane during handover because LTE supports only hard handover,the interruption degrades the delay performance of UDP because no data can be sent between eNB and UE during the interruption . It has also been pointed out that longer interruption times may degrade the TCP throughput performance. If the interruption time exceeds the retransmission timeout of a TCP session, the congestion avoidance algorithm of TCP reduces the TCP throughput. This phenomenon is called spurious timeout. This paper evaluate the TCP throughput and UDP delay by using indoor and outdoors testbeds. The indoor testbed consists of EPC (evolved packet core), eNBs, and UEs (user equipment) in full compliance with 3GPP release 8. A fading simulator is used to emulate the propagation environment between the eNBs and the UEs. The outdoor testbed is located in Kumagaya city in Japan. We evaluate the TCP performance in a real propagation environment. [4][9]

5. The Long Term Evolution of UMTS is just one of the latest steps in an advancing series of mobile telecommunications systems. The 3GPP standard for Home eNodeB, LTE femtocell, which is one of the best approaches to reduce the Operating Expenditures (OPEX) for operators. Femtocells are low- power access points, providing wireless voice and broadband services to customers primarily in the home. With the deployment of the Home eNodeB, the handover between femtocell and 3GPP macrocell networks is become more and more important in the LTE based networks. Thousand of femtocells within a macrocell area will create a large neighbor cell list and interference problem. So the modifications of handover procedures for existing networks are needed. The optimization of handover procedure and algorithm will improve the performance of both the femtocell and LTE networks. [5][10]

Handover Techniques

Handover techniques in wireless communication network can be classified into two main handover techniques types. The first technique type is called Hard Handover (HHO) and the second technique type is called Soft Handover (SHO).

Hard handover The concept of HHO is to break before make. That means the old wireless link connection is broken from the source eNB before a new connection is activated to the target eNB. The UE can communicate with one eNB only in each time slot during HHO. That means, after release the connection from source eNB, the new connection is set up and activated to target eNB. Whereas, after the signal strength from a target eNB exceeds the signal strength from the source eNB the HO start the execution.

Soft handover The SHO handover technique is make-before-break method. That means a new wireless link connection to the target eNB is established while the old connection with source eNB is maintained. The UE simultaneously receive all services data from several active eNBs. Under this technique, there are two main SHO techniques in wireless mobile communication system. The first technique is called Macro diversity Handover (MDHO) and the second technique is called Fast Base Station Switching (FBSS).

Macro Diversity handover (MDHO) In MDHO, a list of BSs is maintained by the MS and BS. This set of BSs is called an Active Set or Diversity Set. Under this technique, MS have the ability to communicate with all BSs in the Active Set as. For DL, MS received and performed data from all the Diversity Set BSs. In the UL, all Diversity Set BSs are received and performed information from MS. Furthermore, the Neighbor BSs can receive the signal from MS, but the signal strength is not sufficient to allow Neighbor BS to be added to the Diversity Set. MDHO supports fast and seamless handover. In addition, MDHO is more stable and gives better performance in term of fast and seamless handover. On the other side, MDHO is more complex for its architecture and during handover procedure than HHO. Therefore, utilizing MDHO will increase system overhead and more network resources will be wasted. This technique is common in UMTS systems and also applied in WiMAX family.

(II) Fast base station switching (FBSS) In FBSS technique, MS and BS maintain a list of access BS called a diversity set and communicates with BS in each frame similarly as in MDHO. The MS continuously monitors the base stations in the Active Set and defines one BS as an "Anchor BS" based on the received signal strength. The Anchor BS is the only BS of the Diversity Set that MS can communicates with for all UL and DL messages including management and traffic connections. This type of handover supports smoother data transition from source to target eNBs and less system overhead than MDHO. In other side, FBSS has high data lost latency and outage probability comparable to MDHO.

Fractional soft handover technique (FSHT) FSHO technique has been proposed in LTE-Advanced system based on Carrier Aggregation (CA) technique. The main concept of FSHO technique is to partially perform soft handover for VoIP service. In this proposed technique, they classify the service to VoIP and non-VoIP services. During the handover procedure, VoIP services are transmitted from both source and target eNBs, while non-VoIP service are transmitted by source or target eNBs.

D. Combined partial resuse and soft handover This scheme proposed an inter-cell interference mitigation scheme based on a combination of partial reuse and soft handover for an OFDMA DL system. The objective of this scheme is to improve the average cell throughput by considering the data rate fairness among the users, compared to the conventional partial reuse scheme, especially at the cell boundary and during handover occurring from source to the target eNBs in LTE- Advanced system. In addition, utilizing this scheme is resulting in a low soft handover overhead.

E. Semi soft handover technique (SSHT) SSHO technique has been proposed utilizing macro diversity method, which permits both HHO and SHO advantages for services over multicarrier-based broadband networks to be retained. This hybrid handover method is known as Site Selection Diversity Transmission (SSDT). This technique represents a possible solution for multicarrier systems. The basic concept of SSDT is to selectively transmit each downlink symbol according to channel quality from each BS.

F. Handover preparation with handover negotiation In the preparation stage source eNB prepares one or more target cells then the source eNB selects the target PCell and the source eNB alos provides the target eNB with the list of cells. The target eNB selects and identifies the SCells that are configured for use after handover. The target eNB use RRC Connection Reconfiguration message, either including the mobility Control Info or not to configure SCells. Once target eNB recives handover command message from the source eNB , the handover negotiation mechanism is executed between two eNBs. The target eNB detects the current identifies SCells once upon reciving the handover negotiation request message. If there are duplications between the identified SCells by target eNB and the current configured SCell by source eNB, target eNB will pre-assign CC for UE,also the target eNB should send handover negotiation request ask message back to source eNB.

G. Synchronous handover Execution When handover negotiation has been executed successfully a dedicated RACH preamble that is allocated for UE by target PCell is send to source PCell with handover command ACK message and transparently forward to UE then handover execution is processed by means of RCC connection reconfiguration including the mobility control info. UE reconfigures PCells and a set of SCells, then attempts to access the target PCell at the first available RACH occasion according to Random Access Resource Selection. Since there are duplication between the reconfigured SCells by UE and original configured SCells by source PCell before handover i.e curren SCells maintain the user data forwarded to UE by source eNB, then the duplicated SCell could provid UE with user data during the RACH access process i.e. the handover is synchronous.

LTE-Advanced System

1. Fast and seamless handover techniques

Fast and seamless handover are important features in wireless systems. Fast handover is a network re-entry procedure with minimum handover latency without any explicit interest in packet loss and interruption time. While, seamless handover are a network re-entry procedure with the capability for UE to contact with the target eNB before initiating a network re-entry control message transaction. Obviously fast and seamless handover depends on the type of user services. For example, the real-time applications such as video conferencing and streaming media are required high data rate and wider bandwidth. So that, there is decreasing for the connection to these real-time applications (i.e. video conferencing and streaming media) will be probably noted to the users during handover from the source to the target eNBs. But in other side, browsing a website or transferring a file not required high data rate compare to the real-time applications, so the user does not have noticed anything during handover process. As a result, the important crucial factors for fast and seamless handover are the latency and packet loss. These two factors have to be as small as possible to make the handover fast and seamless.

2. Supported for legacy handover

The legacy technologies in LTE-Advanced system are refer to Global System for Mobile Communications (GSM), Enhanced Data Rates for GSM Evolution (EDGE), radio access network (GERAN) and Universal Mobile Telecommunications System (UMTS) families: Wideband Code-Division Multiple Access (WCDMA), high-speed data packet access (HSDPA), and HSDPA+ [12]. LTE-Advanced system (R10) specification is required to be backward compatible with GSM family networks. The latest handover techniques in LTE-Advanced system have to support for Legacy GSM network. Since deployment of LTE-Advanced system might be installed over the existing GSM, EDGE, GERAN, UMTS, WCDMA, HSDPA, HSDPA+ and LTE network, the LTE- Advanced UEs are designed to support mobility capability across the coexisting deployment of legacy and advanced networks without critical constraints on service continuation.

3. Lte-Advanced Relay Architecture

Each eNB has a limited serving area. But it is more difficult to deploy a new eNB since residents have much healthy concern about the electromagnetic wave. Thus, the idea of signal relaying by a smaller entity called relay node (RN) has been proposed and becomes an LTE-Advanced specification.There are two different RN types depending on its functionalities. Type 1 RN has the OSI layer 3 functionality that act as a regular eNB for UE and act like an UE for DeNB. Type 2 RN just operates at OSI layer 2 functionality that provides messages decoding and forwarding mechanisms and is transparent to both UE and DeNB. The system management becomes more complicated if the operators deploy different types of RN. Although the architecture and signaling between DeNB and RN are still under discussion and design, RN relaying mechanism already becomes a key technology of LTE-Advanced. How to gain the benefit from RN, fit the QoS requirements and reducing total cost of ownership are the significant research topics. An RN firstly connects to an eNB in order to enlarge eNB's serving area. The eNB that connected with RN is called Donor eNB (DeNB) since it needs to provide more functionality for serving both UE and RN. The interface used between RN and DeNB is called Un. The RN has S1 and X2 interfaces since it act as a regular eNB for UE. Furthermore, an RN is similar to a UE that is running with RRC (Radio Resource Control) and NAS (Non Access Stratum) functionalities.

TCP throughput Evaluation

We evaluate TCP throughput by using two eNBs and three UEs. UE1 is moving between eNB1 and eNB2. UE2 and UE3 are statically attached to eNB1 and eNB2, respectively. Note here that the scheduler in each eNB assigns all available radio resources to the UE(s) by using a proportional fair algorithm. Therefore, the interference from the neighboring cell includes not only reference and control signals, but also the user plane signal if enough user plane data is supplied from the eNBs to the UEs. UE1 (moving UE) is connected to both eNB1 and eNB2. At UE1, RSRP (reference signal receiving power) for eNB1 is controlled by the programmable attenuator and it is varied periodically between -80 and -100dBm.The increase/decrease rate of RSRP is set at 1dBm/sec. On the other hand, RSRP for eNB2 is set to be a constant at -90dBm. At first, UE1 is attached to eNB1 because RSRP for eNB1 is higher than eNB2. As the RSRP for eNB1 decreases and if the RSRP for eNB2 is A dB higher than eNB1 for T milliseconds, where A is A3-offset and T is TTT, handover is triggered and eNB2 becomes the serving eNB for UE1. In this way, UE1 repeatedly hands over between eNB1 and eNB2. Note here that UE2 and UE3 remain attached to eNB1 and eNB2, respectively. FSs (Fading simulators) are inserted between the eNBs and the moving UE to emulate a radio propagation environment. The propagation model used in the indoor experiment is ETU (Extended Typical Urban) and maximum Doppler frequency fD = 70Hz, which corresponds to velocity of 35km/h. A server PC is located at EPC side and client PCs are located at UEs. TCP streams are generated by Server and sent to Client 1, 2 and 3.The congestion avoidance algorithm is BIC (binary increase congestion control) and FRTO (forwarding retransmission timeout) is employed to avoid spurious timeout. To more closely look at the TCP throughput degradation just before handover, we filtered the TCP throughput by measuring RSRP for eNB1 and eNB2 at the same time as the TCP throughput measurement. If the difference between the RSRP for the serving eNB (RSRPserving) and the RSRP for the target eNB (RSRPtarget) is less than a predefined filtering threshold (F), i.e.

RSRPserving - RSRPtarget < F, (1)

The filtering threshold is set at F = 2dB. (Handover was repeated for 60 times to gather many samples.), smaller A3-offset improves TCP throughput. Another interesting point here is that Ping-Pong handovers are observed when A = 0dB. Ping-Pong handovers generate more interruptions and degrade the TCP throughput performance. However, we can obtain better TCP throughput performance with A = 0dB. Therefore we can conclude that a Ping-Pong handover does not significantly affect TCP throughput performance.

UDP delay Evaluation

For UDP traffic transmission rate is set at 1 Mbps. Since we set the UDP packet size at 1250 bytes, a packet is sent to Client 1 every 10ms. Probe1 and Probe2 in Fig. 3 measured one-way delay of UDP packets. When a packet goes through both probes, the transmit times are recorded. Since Probe1 and Probe2 are synchronized by GPS (global positioning system), it is possible to calculate the one way downlink delay by taking the difference between the recorded times at Probe1 and Probe2 for each packet. The one way downlink transmission delay
measured during LTE handover with A = 6dB and T = 320ms
when the RSRP increase/decrease cycle executed to trigger four handover events. The spikes are
created by the interruptions caused by the handovers. The
observed interruption time was around 80ms. Although the
results are not given here, we obtained similar interruption
times for other A and T values in this particular radio
propagation environment and traffic pattern.

LTE Femtocell System Architecture

For the E-UTRAN HeNB architecture, the discussions for the LTE femtocell standards are undergoing in the Femto Forum, in NGMN Alliance and in 3GPP. While the architecture has not been finalized, there is a strong consensus to keep it as flat as possible, following the principles of 'all-IP' networks adopted in the LTE standards. The debate is still going on as to whether there is a need for a signaling aggregation element or whether the evolved packet core (EPC) itself should be able to support femtocells directly, which has a set of S1 interfaces to connect the HeNB to the EPC. With the involved of Home eNB Gateway (HeNB GW), it equivalent to expanding the S1 interface between HeNB and core network, and more HeNB can be deployed. We can assume that the HeNB GW worked at control plane, especially the concentrator of the S1-MME. The HeNB side S1-U interface can be terminated at HeNB GW or the logical connection between HeNB and S-GW by directly user-plane.

To integrate with LTE macrocell networks better, the HeNB GW should appears to the MME as an eNB, the HeNB GW appears to the HeNB as an MME between the HeNB and the Core Network though there may be tens of thousands of femtocells in a traditional LTE macrocell, HeNB GW may also have interface to operator's O&M system for configuration and control.The S1 interface between the HeNB and the EPC is the same whether the HeNB is connected to the EPC via a HeNB GW or not. Here we choose the LTE femtocell system architecture based on concentrator.

The interfaces between the HeNB and the EPC are the standard S1-MME and S1-U, with the HeNB GW optionally providing aggregation function for the S1- MME. The S1-U interface adopts a direct tunnel approach, but optionally also this interface can be aggregated by the HeNB GW. In this case, the HeNB GW may also provide support for user plane multiplexing, for efficient transmissions over limited bandwidth links. The functions supported by the HeNB shall be the same as those supported by an eNB and the procedures run between a HeNB and the EPC shall be the same as those between an eNB and the EPC.

Performance Analysis

Since UE supports only hard handover scheme in LTE- Advanced environment, UE has to disconnect with source RN/eNB and then reconnect with RN/eNB among its handover procedure. By using the forwarding mechanism can reduce the opportunity of packet loss. In order to analyze forwarding cost, we define the transmission cost in each link listed as follows.

HPGW_SGW: the hop count between P-GW and S-GW HSGW_DeNB: the hop count between S-GW and DeNB HInter_DeNB: the hop count between Source DeNB and

Target DeNB
 HRadio_RN: the hop count between DeNB and RN HRadio_UE: the hop count between RN and UE P_Timehandover: the percentage of handover period in

 Costdata_trans: the end-to-end transmission cost in EPC and E-UTRAN

Regular transmission cost

Costdata_trans = P_Timehandover (HPGW_SGW + HSGW_DeNB + 3HRadio_RN + HInter_DeNB) +

( 1 - P_Timehandover) (HPGW_SGW +(HSGW_DeNB + HRadio_RN + HRadio_UE)

Smart forwarding transmission cost

Costdata_trans = P_Timehandover (HPGW_SGW + HSGW_DeNB + HRadio_RN + HInter_DeNB) +

(1 - P_Timehandover) (HPGW_SGW + HSGW DeNB + HRadio RN + HRadio UE)


Hop count value













Based on the transmission cost listed above, we can easily compare the difference between regular transmission cost and the cost with proposed smart forwarding scheme. The comparison of end-to-end transmission cost between regular and proposed schemes shows the comparison of radio access portion only between regular and proposed schemes. The Physical frame structure that has been introduced in [17] is considered in our simulation. Based on that, the cell throughput can be calculated for every sub carriers at the UE side. Where, it is measured in every iteration time during all the simulation time and then taking the average over all the users that are simultaneously active in the cell. The evaluation performance in term of cell throughput is measurement is based on Shannon formula that has been defined in [18]. It can be formulated as

Data a = BW / a x (BWeff x log2 (1+SINRa / SINReff)

Where, BW is the total system bandwidth in Hz, BWeff is the system bandwidth efficiency, SINR0 is the achieved SINR, 0 is frequency reuse factor, it is assumed to be one (0 =1), which means only 1/0th of the spectrum can be used by one cell and SINReff is SINR implementation efficiency and formulated.


1. Both LTE and LTE-Advanced are backward compatible with 3G/UMTS cellular systems and would consider more future prospect and opportunity of becoming 4G standard. The LTE-Advanced adds a new entity called relay node (RN) to enlarge service coverage.Performance analysis shows that the proposed scheme can efficiently reduce forwarding cost as well as wireless channel usage during handover.[6]

2. CA technique is expected to play a pivotal role in future release of LTE Advance system by the helping to enhance network coverage and improve service quality the proposed scheme called combined handover scheme is a handover optimized strategy under CA mode. The proposed approach can reduce much of the interruption time in handover region and it is flexible and easy to implement.[7]

3. The existing handover technique that is utilized in LTE- Advanced system is known as HHO. HHO offers reduce architecture and handover procedure complexities. But on the other hand, there are several limitations when performing HHO, such as high latency, handover procedure unreliability, high outage probability and data lost. Those handover techniques support seamless handover, but suffer from some flaws such as inter-cell interference coordination (ICIC), interference mitigation technologies, latency, unreliability and some data lost during handover. Moreover, implementing FSHO based on CA has been investigated, which result in improve system performance in term of cell throughput in everywhere in the cell and user's handover numbers much better than the system that has been implemented with one component carrier only (Non-CA).[8]

4. We evaluated TCP and UDP performance during LTE handover by using both indoor and outdoor experiments. A3-offset is a key parameter in LTE handover and several values were examined to determine its impact on the performance of TCP and UDP during handover. TCP throughput is improved by using a small A3-offset value because it effectively suppresses interference from neighboring cells. Decreasing the A3-offset value induces more Ping-Pong handovers, but this does not affect TCP throughput significantly. We also confirmed that the interruption time caused by a handover event was around 80ms.[9]

5. Modified signaling procedure of handover is presented in the Home eNodeB gateway based femtocell network architecture. A novel handover mechanism based on the UE's velocity and QoS have been studied. The comparison with the traditional algorithm shows that the algorithms proposed in this paper have a better performance in the rate of unnecessary handovers and the average number of handovers, especially in Medium and High mobile speed, for a small penalty of signaling overhead. In the future, we plan to investigate the quantitative effect of different handover algorithms on the signaling overhead. [10]