Advertisement

Wireless Personal Communications

, Volume 100, Issue 4, pp 1619–1632 | Cite as

Interdependency Analysis of Inter-Site-Distance on Configuration of Handover Control Parameter in LTE-A HetNets

  • A. Saraswathi Priyadharshini
  • P. T. V. Bhuvaneswari
Article

Abstract

Handover (HO) is the procedure performed to maintain the ongoing session of mobile User Equipment (UE). In Long Term Evolution-Advanced (LTE-A) network, the UE-assisted, network controlled hard HO procedure is suggested by 3GPP. The adoption of the traditional HO mechanism in Heterogeneous Network (HetNet) concept of LTE-A imposes several challenges such as increase in frequent and unnecessary handover rate. This leads to complexity in network planning and maintenance resulting in inefficient resource utilization. Further, the increase in radio link and HO failures are caused due to high-speed UEs. To overcome these challenges, the HO control parameters such as Hysteresis Margin, Cell Specific Offsets, Frequency-Specific Offsets, A3offset and Time-To-Trigger need to be properly configured in consideration with environmental, network and UE characteristics. Hence in this research, an attempt has been made to investigate the impact of these control parameters on HO performance in HetNet environment. Three types of analysis that includes (1) the impact of Inter-Site-Distance, (2) the impact of offloading and (3) the impact of velocity are carried out. The HO performance in terms of HO success rate has been analyzed for Macro–Pico and Pico–Macro scenarios. The results have confirmed that the considered control parameters have a high impact on HO performance.

Keywords

Heterogeneous Network Handover Control parameters Inter-Site-Distance User Equipment 

1 Introduction

The Long Term Evolution (LTE) is the cellular standard defined in Release 8 of the Third Generation Partnership Project (3GPP) that aims to focus on Flat-All-IP based architecture [1]. It encompasses the core network called Evolved Packet Core (EPC) and an access network named Evolved-Universal Terrestrial Radio Access Network (E-UTRAN). The E-UTRAN is the network architecture consisting of base stations called eNodeB. The prominent functions associated with the eNodeB are Radio resource management, IP header compression, Location management and Handover (HO) management. The HO is the mechanism which plays an important role in maintaining the active session of the User Equipment (UE) during its mobility. In LTE network, UE-assisted, network controlled hard HO procedure has been suggested by 3GPP. In this, the serving eNodeB configures the UE to transmit Measurement Reports (MR) through Measurement Configuration (MC) message. Upon receiving the MR, the Serving eNodeB evaluates the reports and decides the target eNodeB based on the availability of resources at the neighboring eNodeB. Once the HO command is received from the serving eNodeB, the UE relinquishes the existing communication with it and establishes new connection with the target eNodeB. Concurrently, the occurrence of the above HO procedure information is communicated to the core network.

On the other hand, to meet the surging data traffic demand, Carrier Aggregation, Multiple Input Multiple Output (MIMO) antenna, Relaying and Heterogeneous Network (HetNet) are the technical enhancements incorporated in Release 10 of 3GPP and specified as LTE-Advanced (LTE-A) Network [2]. Among these enhancements, the concept of HetNet plays a vital role in improving the capacity of the network by overlaying small cells onto the layer of Macro eNodeB. Though the HetNet deployment increases the network capacity by offloading UEs from Macro eNodeB, adoption of traditional HO mechanism in this deployment degrades the network performance. It encounters several challenges such as (1) increase in Radio Link Failure (RLF) and HO Failure (HOF) rate, (2) increase in Ping-Pong and unnecessary HOs, (3) occurrence of small cell under utilization due to the massive disparity in transmission power levels of Macro eNodeB & small cell and (4) increase in the complexity of network planning and maintenance [3].

Thus, the HO performance depends on various factors such as (1) the transmission time of MR (MRTT), (2) availability of radio resource at the target eNodeB and (3) the impact of environmental, network and UE characteristics in HetNet environment. The factor MRTT plays a dominant role in improving the HO performance. It depends on the reporting criterion specified in the MC message. The MR can be sent either periodically or based on the occurrence of some pre-defined event. To avoid wastage of network resources, event based reporting is mostly preferred by the network operators. For Intra-Radio access technology HO, six events are defined by 3GPP. Out of these, A3 event [4] is predominantly utilized as it compares the signal strengths of both the serving eNodeB and the target eNodeB to execute the HO mechanism.

However, due to the dynamic nature of the wireless environment, frequent initiation of A3 event might occur. To avoid this, a control parameter called Time-To-Trigger (TTT) has been utilized to verify the persistence of A3 event. The MR is transmitted only when the A3 event occurs and persists for the duration of TTT. The higher value of TTT delays the initiation of HO procedure while a lower value of TTT initiates early transmission of MR. Thus, the transmission time of MR depends on the control parameters involved in A3 event and also the value of TTT. The control parameters that decide the occurrence of A3 event are Hysteresis Margin (HM), A3Offset (A3 offset ), Cell-Specific-Offsets (CSO) and Frequency-Specific-Offsets (FSO). As this work focuses on Intra-RAT and Intra-Frequency based HO procedure, the control parameter FSO is not considered in this analysis. The parameter HM ensures the triggering of A3 event when the received signal power from the target eNodeB becomes sufficiently higher than that of the serving eNodeB. It ranges from 0 to 15 dB. The parameter A3 offset is configured to enhance the current measurement of serving eNodeB in comparison with the neighbor eNodeB. It ranges from − 15 to 15 dB. The parameter CSO is utilized for load balancing and its value ranges from 0 to 16 dB. There are 16 values defined by 3GPP for TTT. Depending on the value of the above mentioned parameters, the transmission time of MR is either advanced or delayed and it influences the success of HO procedure.

From the literature survey made, it has been found that no work which analyzes the combined impact of all the control parameters on HO performance for HetNet scenario under varied environmental and user characteristics, exists. Hence in [5], an attempt has been made to analyze the combined impact of control parameters on HO performance for homogeneous scenario of Macro–Macro (MM) HO. The impact of HM, A3offset, and TTT on HO procedure triggering condition has been analyzed for varied Inter-Site-Distance (ISD). From the simulation results, the following conditions have been inferred: (1) whenever the sum of HM and A3offset value becomes negative, the triggering condition of A3 event is too early which results in incorrect HO procedure, (2) however, when it is positive, the A3 event triggering condition is too late, and this results in call dropping and (3) hence to avoid call dropping, A3 event triggering condition needs to happen exactly in the middle of the overlapping region which is realizable when the sum of HM and A3offset is zero. In this paper, the analysis has been extended to HetNet scenario involving Macro and Pico eNodeB. The impact of ISD, offloading and UE velocity on HO performance has been investigated for Macro–Pico (MP) and Pico–Macro (PM) HO scenario.

The organization of the paper is as follows: Sect. 2 discusses the state of art related to HO control parameter optimization. Section 3 describes the proposed system model and procedures involved in performance analysis. Results and inference of the proposed model are elaborated in Sect. 4. Section 5 concludes the paper with future directions.

2 Literature Survey

This section details the state of art in the configuration of HO control parameters involved in Macro–Pico deployment. In [6], the authors developed adaptive and grouping techniques to analyze the impact of TTT control parameter on MP and MM HO performance in terms of Radio Link Failure (RLF) and Ping-Pong rate. Initially, they examined the HO performance by configuring various TTT values for different UE velocity in both HO types. The adaptive TTT values which resulted in RLF rate of 2% were determined with respect to the HO type. And then, to simplify the configuration of TTT and retain the adaptability, the concept of grouping was introduced in which the active UEs were classified into three ranges based on their speed. The corresponding TTT values for each range of velocity were identified. From the results obtained it was inferred that significant improvement on MP and MM HO performance was realized by adopting adaptive TTT values when compared to the configuration of fixed TTT values.

The analysis presented above has been further enhanced in [7] with the inclusion of HM, which is yet another control parameter related to HO performance. This parameter provides the influence of channel characteristics on HO performance. The work cited analyzed the impact of both HM and TTT on HO performance for Macro–Macro, Macro–Pico, and Pico–Macro scenarios. From the simulation results obtained, it was inferred that adaptively selected TTT resulted in better performance and consideration of HM mitigated the effect of Ping-Pong rate to a greater extent.

In [8], the authors analyzed the impact of TTT on HO performance. They also analyzed the significance of A3offset to optimize the decision of handover. Further, the impact of CSO was analyzed to explore the usage of small cell. It was concluded that the performance of HO executed between macro and small cell was efficient only when the dwell time of UE was prolonged in the small cell, which depended on the mobility pattern of UE. The conclusions arrived from the analysis were as follows: (1) positive value of A3offset delayed triggering of A3 event which resulted in increase in HO failures while negative value of A3offset increased the Ping-Pong rate, (2) longer TTT reduced unnecessary HOs while smaller TTT reduced HO failures but increased Ping-Pong rate, and (3) larger value of CSO increased the Time of Stay (ToS). Hence, optimal selection of control parameters played vital role in improving the HO performance.

The impact of ISD between macro and small cell on HO performance was analyzed in [9]. The HO performance analysis in terms of HO rate, RLF, PP and HO failure rate were carried out for configuring the control parameter TTT. Based on the simulation results obtained, closed form expressions for the aforementioned probabilistic analysis were derived. The results showed that for small ISD, HO success rate was high and vice versa. On the contrary, the RLF and PP rate decreased with larger ISD and vice versa. Similarly, in high mobility scenarios, the HO of UE at high speed needed to be encouraged as it resulted in higher small cell usage. Conversely, in low mobility scenarios, the HO of UE moving at high velocity should be prevented to reduce the small cell overload. It could be concluded that better HO performance could be achieved when the selection of TTT was done based on the ISD, profile of UE and network characteristics.

From the above literature survey, it became clear that the configuration of control parameters plays a significant role in the improvement of HO performance. However, could also be seen that the combined effect of these parameters remained to be explored. Hence, an attempt has been made in this research to study the interdependence of the control parameters and their impact on HO performance.

3 Proposed Handover Performance Analysis

The proposed Handover performance analysis involves three phases which include (1) Evaluation of A3 Event Trigger Point, (2) Determination of Successful HO, and (3) Impact Analysis as shown in Fig. 1. The system model of the considered scenario is detailed in this section. Let MeNB1 and MeNB2 be the serving and target Macro eNodeB (MeNB) placed with an Inter-Site-Distance of ISDMM (Macro–Macro). Let PeNB be the Pico eNodeB placed between MeNB1 and MeNB2 which results in two different ISDs, namely ISDMP (Macro–Pico) and ISDPM (Pico–Macro).
Fig. 1

Stages involved in the proposed analysis

3.1 Evaluation of A3 Event Trigger Point

In the proposed analysis, the value of FSO is assumed to be zero as the considered HO scenario is Intra-Frequency HO. In this phase, the value of A3 event trigger point (HOA3) is evaluated. It represents the distance at which triggering of A3 event occurs due to the mobility of UE and it depends on the HO control parameters considered and their associated values.

The parameter CSO is utilized in the HO scenario whenever the target eNodeB happens to be the small cell. Further, the value of CSO is kept zero for MeNB as it operates with maximum transmit power and it takes values between 0 and 16 dB for PeNB. Hence, the value of CSO is included only in MP HO scenario. The value of HOA3 for MP HO is computed using Eq. (1) for UE mobility from MeNB to PeNB. It involves control parameters such as HM, A3offset, and CSO. Similarly, for UE mobility from PeNB to MeNB, the value of HOA3 for PM HO is computed using Eq. (2). It involves only two control parameters, namely HM and A3offset.
$${\text{HO}}_{{{\text{A}}3}} = M_{T} + CSO{-}HM > M_{S} + A3_{offset}$$
(1)
$${\text{HO}}_{{{\text{A}}3}} = M_{T} {-}HM > M_{S} + A3_{offset}$$
(2)
where MS and MT are the Reference Signal Received Power (RSRP) of serving and target eNodeB measured in dBm. The ranges of HM and A3 offset values considered in this research are as per 3GPP specification [4]. In the proposed analysis, the values of both HM and A3offset are taken in the intervals of 5 dB (HM = {0, 5, 10, 15 dB} and A3offset = {− 15, − 10, − 5, 0, 5, 10, 15 dB}) and the values of CSO are in the intervals of 4 dB {0, 4, 8, 12, 16 dB}. Based on these values 140 combinations of control parameter values results for MP HO and as the impact of CSO is not utilized in PM HO scenario, it results in 28 combinations. Additionally, the values of TTT as specified in 3GPP are considered for both the scenarios. The combinations that produce the same value of HOA3 are grouped.

3.2 Determination of Successful HO

After determining the occurrence of A3 event entering condition from the previous stage, the persistence of A3 event entering condition is verified for the duration of the TTT values specified in Table 1 to avoid frequent and unnecessary HO. Further, three conditions that make the triggered HO successful are validated. They are (1) successful reception of UE generated MR by serving eNodeB, (2) successful reception of HO command message by UE after HO preparation time (50 ms as per 3GPP), and (3) persistence of radio link with the target eNodeB after HO execution time (40 ms as per 3GPP). In the simulation environment, the above conditions are arrived in terms of the Uplink (UL) and Downlink (DL) RSRP. The above mentioned three conditions are represented by the following equations:
$${1. \quad \text{UL}}_{\text{RSRP}} {\text{with respect to MeNB}}_{1} \ge {\text{RSRP}}_{\hbox{min} }$$
(3)
$${2. \quad \text{DL}}_{\text{RSRP}} \,{\text{of MeNB}}_{1} > - 130\,{\text{dBm}}$$
(4)
$${3. \quad \text{UL}}_{\text{RSRP}} {\text{with respect to MeNB}}_{2} \ge {\text{RSRP}}_{\hbox{min} }$$
(5)
where RSRPmin is the minimum required RSRP by the UE to camp on respective eNodeB while the values of ULRSRP and DLRSRP are computed using the Eqs. (6) and (7)
$${\text{UL}}_{\text{RSRP}} = {\text{TR}}_{\text{UE}} - PL_{UE - MeNB} - G_{T} - G_{R}$$
(6)
$${\text{DL}}_{\text{RSRP}} = TR_{MeNB} - PL_{MeNB - UE} - G_{T} - G_{R}$$
(7)
where
  • TRUE and TRMeNB are the transmitting power of UE and MeNB

  • PLUE-MeNB is the path loss between UE and MeNB

  • GT and GR are the transmitter and the receiver antenna gains

Table 1

Simulation parameters

Parameter

Configuration

Macro and Pico eNodeB Transmitter power

46 and 30 dBm

Macro propagation model

128.31 + 37.06 (log(R)), where R in Km

Pico propagation model

140.7 + 36.7 (log(R)), where R in Km

UE transmit power

30 dBm

Time-to-trigger

{0, 40, 64, 80,100, 128, 160, 256, 320, 480, 512, 640, 1024, 1280, 2560, 5120 ms}

Cell-specific-offset

0, 4, 8, 12, 16 dB

HO preparation time

50 ms

HO execution time

40 ms

Velocity of UE (V)

Urban: Up to 110 km/h

Radio link failure

RSRP < − 130 dBm

Minimum required RSRP (RSRPmin)

− 110 dBm (to camp on respective eNodeB)

eNodeB antenna gain after cable loss

15 dBi

UE antenna gain

0 dBi

The combination of control parameters which satisfies these three conditions is referred to as HO success groups. Also, the maximum applicable value of TTT is identified for each group.

3.3 Impact Analysis

This section describes the scenario considered in the proposed HO performance analysis. The impact of control parameters on HO performance is analyzed in the urban environment with three different cases. First, the optimal deployment of PeNB to improve HO performance is analyzed. It involves three different scenarios with fixed ISDMM which are as follows: (1) PeNB placed near MeNB1 as shown in Fig. 2a where ISDMP < ISDPM, (2) PeNB placed in the middle of MeNB1 and MeNB2 as shown in Fig. 2b where ISDMP = ISDPM, and (3) PeNB near MeNB2 as shown in Fig. 2c where ISDMP > ISDPM. For a fixed value of ISDMM and ISDMP, different control parameter combinations that result in HO success are identified. The analysis is further extended to different values of ISDMM ranging from 500 m to 800 m and for Macro–Pico and Pico–Macro HO scenarios.
Fig. 2

a Position of Pico eNodeB Near Serving Macro. b Position of Pico eNodeB in middle of Serving and Target Macro. c Position of Pico eNodeB Near Target Macro

In the second case, the impact of CSO on HO performance is analyzed. Figure 3 shows the pictorial representation of virtual cell range expansion by adding CSO varying from 0 to 16 dB.
Fig. 3

Pico eNodeB Range Expansion

Finally, the impact of TTT on HO performance is analyzed for different velocities of UE by keeping constant the remaining control parameters such as CSO, HM, A3offset, and ISDs.

4 Results and Discussion

The simulation and HO performance analysis for the proposed Heterogeneous scenario are discussed in this section. Table 1 presents the values of the parameters utilized in the simulation. The simulator used is MATLAB R2014a.

4.1 Impact of ISD Analysis

In this analysis, both Macro–Pico and Pico–Macro HO performances are investigated for different values of ISDMM, ISDMP, and ISDPM as mentioned in Table 2.
Table 2

ISDs considered for analysis

ISDMM (m)

ISDMP (m)

HO type

Pico near to serving (m)

Pico in middle (m)

Pico near to target (m)

500

100

250

400

Macro–Pico and Pico–Macro

600

100

300

500

700

100

350

600

800

100

400

700

4.1.1 Case 1: Pico–Macro HO

The control parameters involved in this analysis includes (1) HM, A3offset of 28 combinations to determine the occurrence of A3 event, and (2) TTT of 16 values to decide the initiation of HO. Initially, the occurrence of A3 event is verified for fixed values of ISDs and maximum UE velocity of 110 km/h. It has been observed that among 28 combinations, there are certain combinations which trigger the occurrence of A3 event concurrently. Hence, the combinations which trigger the simultaneous occurrence of an A3 event are grouped. It resulted in ten groups from G1 to G10 as mentioned in Table 3. Among these groups, G4 triggers the occurrence of A3 event whenever the received power of target eNodeB becomes equal to that of the serving eNodeB. And, the distance at which A3 event occurs increases when configured with groups from G1 to G10. It indicates that group G1 triggers HO earlier while G10 triggers HO very late.
Table 3

Groups with same HOA3

Groups

(HM, A3offset) in dB

G1

(0, − 15)

G2

(5, − 15) (0, − 10)

G3

(10, − 15) (5, − 10) (0, − 5)

G4

(15, − 15) (10, − 10) (5, − 5) (0, 0)

G5

(15, − 10) (10, − 5) (5, 0) (0, 5)

G6

(15, − 5) (10, 0) (5, 5) (0, 10)

G7

(15, 0) (0, 15) (10, 5) (5, 10)

G8

(5, 15) (10, 10) (15, 5)

G9

(15, 10) (10, 15)

G10

(15, 15)

And then the maximum configurable value of TTT which results in HO success for the corresponding groups is determined. For UE with maximum velocity, TTT value of 80 ms achieves HO success. Figure 4 represents the groups which result in HO success and their corresponding distance of A3 event trigger point from MeNB1 for ISDMM of 500 m and the corresponding positions of PeNB as mentioned in Table 2. It has been inferred from the graph that, groups from G1 to G10 favor HO success when ISDMP > ISDPM. And the number of groups resulting in HO success decreases when ISDMP < ISDPM and when ISDMP = ISDPM. Also, the same group triggers A3 event at different distances for different ISDMP. It shows the dependency of ISDMP in HO control parameter configuration.
Fig. 4

Pico–Macro HO with Macro–Macro ISD = 500 m

To analyze the impact of ISDMM, the above analysis is extended to all the values of ISDMM such as 600, 700, and 800 m along with their respective ISDMP as mentioned in Table 2. The outcome of the analysis is presented in Fig. 5. From the results obtained, it has been inferred that groups G1–G10 result in HO success, when ISDMP > ISDPM for all the ISDMM considered, while for the scenario of ISDMP < ISDPM and ISDMP = ISDPM, the number of groups which result in HO success decreases as ISDMM increases. It has also been observed that, when PeNB is away from MeNB2, HO has to be triggered early. And, it has to be triggered late when PeNB is near MeNB2. It shows that Pico–Macro HO performance can be improved only when the control parameter configuration is made depending on both ISDMM and ISDMP.
Fig. 5

Pico–Macro HO with different Macro–Macro ISDs

4.1.2 Case 2: Macro–Pico HO

The control parameters involved in Macro–Pico HO are HM and A3offset of 28 combinations, CSO of 16 dB, and TTT of 16 values. Hence, the occurrence of the A3 event depends on HM, A3offset and CSO. There are ten groups from G1 to G10 formed as mentioned in Table 2 for CSO of 16 dB. The maximum TTT value that can be applied to the UE traveling at the maximum velocity of 110 km/h is 80 ms. Similar to the analysis carried out above, the occurrence of A3 event for the groups from G1 to G10 with fixed ISDMM and respective ISDMP is analyzed and represented in Fig. 6. The value of HOA3 is with respect to Serving PeNB in meters.
Fig. 6

Macro–Pico HO with Macro–Macro ISD = 500 m

From the simulation results, it has been observed that configuring groups from G1 to G6 result in MP HO success with ISDMM of 500 m for three cases ISDMP < ISDPM, ISDMP = ISDPM, and ISDMP > ISDPM, while configuring groups from G7 to G9 trigger MP HO success for ISDMP = ISDPM and ISDMP > ISDPM. And configuring the group G10 initiates MP HO success when only ISDMP > ISDPM. It also shows that the A3 event occurs at different distances for the same group depending on the value of ISDMP.

Further, to study the impact of ISDMM on MP HO performance, the above analysis has been extended to different ISDMM and represented in Fig. 7. The results indicate that as ISDMP increases, the number of groups which result in MP HO success increases and vice versa. This is because the influence of MeNB1 will be less as long as the PeNB is away from it. Hence the possibility of MP HO success increases. And, not much wide variation is seen with respect to ISDMM. Hence, it can be concluded that the control parameter configuration in MP HO scenario shows higher dependency on ISDMP than on ISDMM.
Fig. 7

Macro–Pico HO with different Macro–Macro ISDs

Overall, it is seen that groups G1 to G6 result in MP and PM HO success irrespective of ISDMM, ISDMP, and ISDPM. The groups from G7 to G10 result in both MP and PM HO success when ISDMP > ISDPM. Hence, the control parameters HM, A3Offset need to be configured depending on the type of HO (i.e., MP or PM) and ISD.

4.2 Impact of Cell-Specific Offset Analysis

The impact of CSO on MP HO performance is analyzed in this section. The value of ISDMM is configured to be 500 m. The PeNB is placed in three different positions as mentioned in the previous analysis. The values of CSO are as per 3GPP specification. It is varied in interval of 4 i.e., 0, 4, 8, 12, 16 dB. For each value of CSO, ten groups resulting in HO success are identified. Collectively, 50 groups are identified for the above values of CSO. From the obtained results shown in Fig. 8, it is inferred that as the value of CSO decreases the number of groups offering HO success also decreases. This is because the higher value of CSO virtually increases the coverage of PeNB when compared to lower values of CSO. This causes both late and early triggered HO to result in HO success. Thus, a higher value of CSO initiates more number of groups to result in HO success and vice versa.
Fig. 8

Impact of CSO on HO performance

Further, the HO success also depends on the position of PeNB. Among the 50 groups considered for MP HO scenario, 22 groups result in HO success when PeNB is deployed near MeNB1, 36 groups in middle position, and 44 groups when positioned near MeNB2. The number of HO success groups differs with respect to ISD for the same value of CSO. Thus from the results, it is concluded that the HO success depends on both ISD and the CSO.

4.3 Impact of Mobility Analysis

The impact of TTT on HO success is analyzed by setting fixed values to parameters such as ISDMM, ISDMP, ISDPM, and CSO. The velocity of UE is varied from 10 to 110 km/h in the interval of 10 km/h. The maximum value of TTT that achieves HO success is determined. The result obtained is depicted in Fig. 9.
Fig. 9

Maximum TTT value versus UE Velocity

From the simulation results, it has been inferred that the maximum value of TTT which determines HO success decreases with increase in UE velocity. This is because higher velocity UEs crossing the small cell experiences less time of stay, whereas, the low mobility dwell for a longer period. Thus, verifying the persistence of A3 event entering condition for the maximum duration of TTT for high-speed UEs results in HO failure. Hence, the value of TTT should be kept lower for high-velocity UEs and vice versa.

5 Conclusion and Future Work

In this research, the combined impact of control parameters such as HM, A3offset, CSO, and TTT on HO performance in HetNet scenario was analyzed. Three different analyses were carried out, namely, (1) Impact of ISD, (2) Impact of Offloading, and (3) Impact of Mobility. The following conclusions were arrived from the ISD analysis: the groups G1 to G6 result in MP and PM HO success irrespective of ISDMM, ISDMP, and ISDPM, while the groups from G7 to G10 result in both MP and PM HO success when ISDMP > ISDPM. Though the values of these groups fall in the range specified by 3GPP, it triggers HO failure for other cases when configured as a combination. The CSO analysis implies that HO success depends on both ISD and CSO. Hence, HO needs to be triggered late for higher bias by configuring late triggering groups and vice versa. The results of mobility analysis have revealed that HO failure can be prevented by configuring the lesser value of TTT when high-speed moving UEs are configured with a smaller value of TTT and vice versa. The results emphasize that the configuration of control parameters strongly depends on various factors such as ISDMM, ISDMP, ISDPM, CSO, and UE velocity. Hence, the control parameters HM, A3offset, and CSO need to be configured based on the value of ISD, while the value of TTT needs to be configured based on UE velocity.

However, the presented analysis considered the linear movement of UE and neglected Quality of Service (QoS) characteristics of UEs. Hence, the random mobility of UEs along with their QoS profile is a possible extension of the presented work. And, a reinforcement learning based algorithm has to be formulated with the conclusions drawn from this analysis. Also, the performance metrics such as a number of handovers, ping-pong handovers, unnecessary handovers, radio link, and HO failures may be analyzed in detail.

References

  1. 1.
    Kottkamp, M., Roessler, A., & Schlienz, J. (2012). LTE-advanced technology introduction. White paper—Rohde & Schwarz (pp. 1–41).Google Scholar
  2. 2.
    Parkvall, S., Furuskar, A., & Dahlman, E. (2011). Evolution of LTE toward IMT-advanced. IEEE Communications Magazine, 49(2), 84–91.  https://doi.org/10.1109/MCOM.2011.5706315.CrossRefGoogle Scholar
  3. 3.
    Contains, I., & Hedlund, A. (2013). HetNets: Opportunities and challenges. White paper—An Ascom Network Testing (pp. 1–22).Google Scholar
  4. 4.
    3GPP TS 36.331, V13.1.0. (2016). Evolved Universal Terrestrial Radio Access (E-UTRA); Radio Resource Control (RRC); Protocol specification. (Release 13).Google Scholar
  5. 5.
    Saraswathi Priyadharshini, A., & Bhuvaneswari, P. T. V. (2016). A study on handover parameter optimization in LTE-A networks. In IEEE First international conference on Microelectronics, Computing and Communication (MicroCom), West Bengal (pp. 1–5). http://dx.doi.org/10.1109/microcom.2016.7522429.
  6. 6.
    Lee, Y., Shin, B., Lim, J., & Hong, D. (2010). Effects of time-to-trigger parameter on handover performance in SON-based LTE systems. In IEEE-16th Asia-Pacific conference on communications (APCC) (pp. 492–496).Google Scholar
  7. 7.
    Lim, J., & Hong, D. (2013). Mobility and handover management for heterogeneous networks in LTE-advanced. Journal of Wireless personal communications, 72, 2901–2912.CrossRefGoogle Scholar
  8. 8.
    Mehta, M., Akhtar, N., & Karandikar, A. (2015). Impact of handover parameters on mobility performance in LTE HetNets. In Twenty-first national conference on communications (NCC) (pp. 1–6). http://dx.doi.org/10.1109/ncc.2015.7084910.
  9. 9.
    Kollias, G., Adelantado, F., & Verikoukis, C. (2015). The impact of inter-site distance and time-to-trigger on handover performance in LTE-A HetNets. In IEEE ICCInternational conference on communications (ICC) (pp. 3969–3974). http://dx.doi.org/10.1109/icc.2015.7248944.

Copyright information

© Springer Science+Business Media, LLC, part of Springer Nature 2018

Authors and Affiliations

  • A. Saraswathi Priyadharshini
    • 1
  • P. T. V. Bhuvaneswari
    • 1
  1. 1.Department of Electronics Engineering, Madras Institute of TechnologyAnna UniversityChennaiIndia

Personalised recommendations