Handover Problems and Solutions for 5G Communication Technology

Introduction

The continuously increasing demand of wireless communications impels the fast development of the next generation wireless communication technology. The 5G communication technology, which scheduled to be realized in the year beyond 2020, is still on its researching stage. Aim at the novel handover problems in ultra-dense network deployments and high-band communications in 5G scenarios.

Handover is an important content of radio resource management. The processing and optimization of handover has a great effect on improving the effectiveness and reliability of the whole system, which plays an important role in modern wireless communication. In 5G mobile communication system, it is necessary to provide high bandwidth and high transmission rate service for different kinds of terminals in several scenarios. Handover technology is an important guarantee of communication continuity and service quality, and it is very important to the overall performance of the network.

We aim to understand the possible solutions to solve the handoff challenge in hybrid 5G. As we move toward 5G, environment becomes so complex that the handoff problem faces with new challenges. 5G will not completely replace the existing technologies but be more integrative and hybrid, combining with existing technologies to provide high-rate and seamless communication service. The data rate in 5G is expected to be roughly 1,000x compared with current technology, hence the handoff problem requires a faster processing. Furthermore, as the number of Base Stations and mobile devices increases, the centralized control may not be efficient. On the contrary, more intelligent mobile devices can play important roles in handoff. Moreover, increasingly serious data security problem reminds users do not share their private information with others. We aim to understand a user centered handoff scheme for hybrid 5G environments and we aim to analyze the proposed solutions and its challenges

Handover in Mobile Communication Systems

A handover or handoff is a process in mobile communications and telecommunication systems in which a connected cellular call or a data session is transferred from one base station (cell site) to another without the session getting disconnected. Cellular services are based on mobility management and handover, allowing the user to be moved from one cell site range to another or to be switched to the nearest cell site for better performance. Handoff can also be described as a process in which a link is transferred from one base station to another base station due to lack of signal strength. Several operations are included in the handoff process such as initiating the handoff, allocating the channel, breaking the connection with the old base station, choosing the new base station.

Types of Handoff

Handoffs can be classified in two categories – hard and soft handoffs. A hard handoff can be thought of as a break before make connection. The base station transfers the user’s call to another cell and the drops the call. In the case of hard handoff, the users link to the prior base station is terminated before or as the user is transferred to the base station of the new cell. In any
case the user cannot be linked to more than one base station at any given time. Hard handoff is used in frequency division multiple access and time division multiple access where adjacent channels use different frequency ranges in order to minimize channel interference.

In the case of soft handoff, a connection to the new base station is made before the connection to the previous base station is disconnected. It is performed through the parallel use of source and destination channels over a period of time. Soft handovers allow parallel connection between three or more channels to provide better service. Soft handoff is effective in poor coverage areas.

Different phases in Handover

1. Handoff Initiation Phase: This phase decides the requirement of handoff. It triggers handoff on the basis of information collected about network, mobile node and user preferences from different layers’ likes’ network layer, transport layer and application layer. These layers provide the information such as signal strength, bandwidth, link speed, throughput, jitter, cost, power, user preferences and network subscription etc. Based on this information, handoff is initiated at appropriate time.
2. Handoff Decision Phase: After initiating handoff in first phase, the decision making phase decides the optimal networks for handoff. The comparison is made between the current and neighboring networks based on parameters such as QoS, signal strength, velocity, direction, cost, etc. There are many methods proposed to take decision about networks such as MADM approaches, fuzzy logic, genetic algorithms, for deciding the destination network. The decision making phase chooses the optimal network for transfer but the actual link transfer takes place in the next phase
3. Handoff Execution Phase: After the selection is made and decision about the target network is taken, link transfer takes place in this phase in which the existing link is re-routed to the new network in a seamless manner. This phase also includes the reauthentication, re-association and re-authorization, and the transfer of user’s context information.

Three strategies are used to detect the need of handoffs

1. Mobile-controlled-handoff (MCHO): The mobile node keeps on monitoring the signals of the current and neighboring base stations and triggers handoff as and when needed.
2. Network-controlled-handoff (NCHO): The base stations monitor the signal between user and the network. It initiates handoff process when there is a drop in the signal strength as the user moves away from the base station.
3. Mobile-assisted-handoff (MAHO): The mobile node assists the network to measure the signal strength between the user and the base station. The networks make the handoff decision based on reports from the mobile node.

Delays in Handoffs

The complete handoff process is carried out by co-ordination and co-operating entities of different layers of OSI model. However, this leads to a system with delay. When data has to be passed between different layers’ different kinds of delays are introduced which hinder in quality of service in handoffs.

1. Link layer delay: This layer is responsible for horizontal handoffs. Before the mobile node switches to neighboring base station, scanning is done. During scanning, mobile node scans all the nearby base stations. It takes some time and sometimes mobile disconnects and then connects to new base station. This kind of delay which is introduced during scanning and call setup with new base station is link layer delay
2. Network layer delay: This layer is responsible for vertical handoff which involves obtaining the IP address of new network and then switching the call from one base station to another which may cause more delay. Thus this kind of delay is called network delay.
3. Transport Layer delay: Many protocols sit at this layer to monitor and manage transport layer handoff. Most important protocol in this case is SCTP which is responsible for multistreaming during handoff. SCTP makes new connection with the neighbor network/base station while maintaining connection with the old. As new connection becomes stable, it breaks the connection with the old base station. This disconnection can be delayed due to reauthentication or re-association phase.
4. Application layer delay: This delay occurs when certain properties of application layer get modified e.g. change in IP using Session initiation protocol.

User Centered Handoff Scheme for 5G Environments

5G will not replace all the current technologies completely but will try to combine with existing technologies to provide higher data rates and seamless communication services. As 5G technologies come into picture, the environment starts becoming more complex and the handoff problem faces new challenges. Data rates in 5G are expected to be higher than the current 4G technology, and as the number of base stations and mobile devices increase dramatically, the use of centralized control may not be efficient. On the contrary, more intelligent mobile devices can play important roles in handoff. Furthermore, increasing data security issues leads to users not sharing their private information with others. Therefore, we want a fast, distributed, privacy preservation and a user centered handoff scheme in hybrid 5G environments.

Consider a scenario in the figure below in which a hybrid 5G environment is constructed using 3G, LTE, WIMAX and 5G base stations. In this scenario, users may need to transfer their network connections from one base station to another base station. This is the handoff problem wherein a user has several available base stations and the user needs to decide to the base station to which the network connection should be transferred. In the figure below as the user moves far away from the 3G base station, the signal strength received from that base station reduces drastically that the user has his network connection to a new base station. The user has 3 choices to which he can transfer his connection to, they are the LTE, WIMAX and 5G base stations. The user has to decide which base station is to be selected.

In this case the handoff problem appears straightforward task of the user selecting the best performing base station. However, it is difficult for the user to know the network selection choices being made by other users in the network. If the same base station is selected by many users, there are high chances of the user being blocked by the network. Therefore, the objectives of network selection are to select a high performance base station and avoid being blocked.

The handoff problem in general can be solved by the network centered approach and the user centered approach. In the network centered approach, networks are responsible for computing and making the decisions while in the user centered approach, users will be in charge of the network selection. To satisfy the requirement of privacy-preservation in hybrid 5G environment, users are not suggested to send their private information out.

User Centered Handoff Scheme for Hybrid 5G Environments

      Figure 1

With this limitation networks can’t obtain adequate information from the users for the network selection. Thus the user centered approach is more suitable for the hybrid 5G environments than the network centered approach. Users are divided into two classes: non handoff users and handoff users.  Non-handoff users will stay in the connections with their current base stations while handoff users will transfer their network connections to new base stations based on limited local information. Local information is the private information of the user which has parameters of the base stations and two parameters related to public information (i.e., the total numbers of handoff and non-handoff users inside each available base station). When a user has to select a new base station in a handoff, the user will calculate the achievable data receiving rates of all the available base stations. Additionally, the user has to infer the network selection behaviors of all the other users to estimate its block probability for each available base station. By jointly considering the achievable data receiving rate and the block probability, the user can select the most appropriate base station in a handoff. 

Relation between Users

Based on limited local information, each handoff user tries to select a new base station which can provide the maximum achievable data receiving rate and minimal block probability. The block probability of a handoff user relates to the network selection behavior of other handoff users and a user has no idea of other handoff users due to privacy preservation. Block probability calculation relies on inference made by a handoff user to another handoff user. To assist a handoff user in inferring the network selection behaviors of other handoff users, it is important to understand the relation between any two handoff users based on their set of available base stations. The relation between a pair of handoff users can be divided into independent relation and correlated relation. When the two users are independent, the network selection behavior of user 1 has no direct impact on user 2 and vice versa. Therefore, in the handoff process a user needs to consider only those users with whom he has some amount of correlation. Correlated relation refers to the scenario when there is atleast one base station available to both the users. To understand the extent to which the two users are correlated, a metric named correlation degree is defined. The correlation degree of two users is the probability of selecting a base station that is available to both from a set of base stations that are available to each user.

     Figure 2

Figure above describes the calculation of correlation degree L(ui,uj) where ui and uj are two users and Bi(t) and Bj(t) are the set of base stations available to each user. x is the set of base stations that are available to both. The value of x is zero if both the users are independent and the correlation degree is zero in this case. If the coreelation degree is 1, all the base stations are available to both the two users.

Network Initiated Handover

In Network Initiated Handover the user gets handover information from Base Station 1. The user then performs quality measurements on Beam Reference Signals (BRSs) from neighboring cells. The user then sends the quality measurement report back to Base station 1. As soon as the Base Station 1 receives the information from the user, it replies with handover command containing connection re-configuration along with target cell Base Station 2. Using the command user then completes handover with target cell base station 2.

User Initiated Handover

Initially the user receives measurement configuration from serving cell base station, it performs quality measurements on BRSs from neighboring cells. The user sends measurement report to serving cell base stations and the base station responds with list of possible neighboring cells to complete the handover. The Base station then provides measurement configuration for user to complete the handover with target base station 2.  The user then completes handover with target cell base station 2.

Figure 3

Ultra Dense Networks

Ultra-Dense Networks (UDN) are networks where there the number of cells is greater than active users. In other words, the density of access points is greater than the density of users. Different quantitative definitions exist for Ultra dense networks exist such as when the measure of cell density is greater than 1000 cells / Km squared or when there are 600 active users / Km squared. Small cells in Ultra Dense Networks are classified into fully functioning base stations which include picocells and femto-cells. It also includes macro extension access points which includes relays and Remote Radio Heads (RRHs)

Ultra Dense Network with Small cells

Figure 4

Small Cells in 5G Technology

Small cells are short range, low power base stations which cover a small geographical area or indoor/outdoor applications. Small cells do have all the basic characteristics of a conventional base station and it is also efficient in handling high data rate for individual users. In advanced LTE and 5G deployments, small cells will play a crucial role in the efficient delivery of high speed mobile broadband and other low latency applications. Small cells are further divided into three classes, they are femtocells, picocells and microcells. Small cells are classified into these three classes based on the number of users the network can handle and the coverage area.

Femtocells – These are small mobile base stations which provide extended coverage for residential applications. They have a coverage area of about 10 to 50 meters. Femtocells are low cost, can support upto 16 users and primarily used for indoor applications.

Picocells – They are another category of small cells which provide extended coverage and are primarily used for small enterprise applications. They have a coverage area of 100 to 250 meters and can support upto 64 users. Picocells are low cost and require around 250 milliwatts of power.

Microcells – In comparison to femtocells and picocells, microcells support a slightly larger number of users. They are capable of covering larger cell size and are used for applications such as smart cities, smart metro due to their high transmission power.

As conventional mobile networks have limitations for further enhancements due to technology, infrastructure and bandwidth, small cells are used with advanced technologies like massive MIMO and beamforming to increase spectrum efficiency and data rate. Small cells work similar to conventional cells with the use of advanced techniques like MIMO, beamforming and millimeter waves for transmission. Small cell enables the deployment of low power transmitting stations. Small cell hardware units reduce complexity making the implementation faster and easier. Small base stations (transceivers) can be fixed on a wall for indoor applications and small towers. Backhaul connections can be made using fiber connections, wired connection and via microwave links. Configuration is less complex and it just needs to be connected to the power source and backhaul. Small cells are primarily added to increase capacity in hot spots with high user demand and to fill in areas not covered by the macro network – both outdoors and indoors. They also improve network performance and service quality by offloading from the large macro-cells. Frequency, power and antenna techniques are some of the factors that affect the cell coverage and data capacity. Small cells can be used extensively in future applications which are given as follows:

1. Disaster management applications and drones
2. Support mission critical services that require low latency and highly reliable network.
3. Internet of thing applications like smart city, smart home and smart healthcare.
4. Support huge number of users during special events like sports and games with multiple cell deployments.

Since 5G technology is a future framework to support various applications, it has to meet all necessary specifications. Small cell concept is an apt solution for delivering enhanced mobile broadband, low latency and reliable service to users. Higher order modulation techniques, MIMO technology and millimeter wave spectrum will ensure proper working of future small cell deployments.

Handover Management in small cells

Several factors need to be accounted while designing handover mechanisms between small cells. The varying link quality of small cells implies that handover between small cells should involve low amount of signaling. Due to the small cell sizes frequent handovers may occur which could contribute in a major way to the signaling load on the network. Load on the core network can be minimized if handovers are anchored locally whenever possible. Additionally, the access control features at small cells mean that the admission control decisions at the target small cell during handover should account for the access privileges available for the mobile user and must not impact the performance of the users already being served at the target small cell.

Sectoring

Sectoring is the process of dividing a congested cell into smaller cells each having its own base station and their antenna height being reduced. Cell sectoring leads to an increase in the channel capacity. Omni directional antennas at the base stations are replaced by several directional antennas. Cell sectoring is done to reduce co-channel interference.

Figure 4.

Cell Splitting with Umbrella Approach

The cell shape is chosen to be hexagonal. This is due to the fact that it successfully avoids overlap that would have occurred between circular shaped cells. The Umbrella approach saves one from building multiple base stations. In the Umbrella cell approach, a single base tower serves all the split cells via antennas which are mounted on it at different heights. Cell splitting leads to increased number of antennas per base station and an increase in the number of handoffs.

Figure 5.

Skipping

With the use of small cells, the number of handoffs required increases in comparison to conventional networks. To overcome this problem, a skipping scheme can be used to reduce the unnecessary handovers.

Figure 6. from Reference [3]

Types of Skipping techniques:

1. Alternating Handover^(a) – This is the simplest of all the skipping schemes. If a user trajectory is passing through lot of small cells it skips every alternate cell. This nearly reduces the number of handovers required to half.
2. Location-Aware Handover Skipping^(b) – The location aware handover skipping scheme looks for the shortest distance between the user trajectory and the target base station to decide the handover skipping. The handover skips associating to the target base station if and only if the minimum distance along the user trajectory and the target base station becomes larger than a pre-defined threshold L. This threshold L can be designed such that the user can skip the base stations in which the trajectory passes through the cell edge only.
3. Cell-Size Aware Handover Skipping^(c) – Cell-size aware handover skipping scheme enables the users to skip handover to target base stations that have a footprint less than a pre-defined threshold S. Since the cell dwell time is dependent on the footprint size of the base station, size aware skipping scheme aims at avoiding large blackout durations. Hence, it allows users to skip small sized cells and associate with large sized cells.
4. Hybrid Handover Skipping: The true cell dwell time is not correctly reflected by the location aware skipping nor the size aware skipping. Hence, combining both schemes gives a better inference about the cell dwell time, which can improve the handover skipping decisions and performance. Subsequently, the hybrid HO skipping scheme combines both location awareness and cell-size awareness to decide which base stations are to be skipped. That is, it takes user location and cell area into account while making the decision for the handoff.

Conclusions

We reviewed a user centered approach for handover in hybrid 5G environment and understood that a user centered approach is needed for 5G technology. We also reviewed 5G cell structure and its impact on handover process.

References

1. Li Qiang, Jie Li, Corinne Touati “A User Centered Multi-Objective Handoff Scheme for Hybrid 5G Environments”, IEEE Commun. Mag., vol. 5, no., 3 July-Sept. 1 2017.
2. R. Arshad, H. Elsawy, S. Sorour, T. Y. Al-Naffouri and M. Alouini, “Handover Management in 5G and Beyond: A Topology Aware Skipping Approach,” in IEEE Access, vol. 4, pp. 9073-9081, 2016. doi: 10.1109/ACCESS.2016.2642538
3. Nisha Panwar, Shantanu Sharma, Awadhesh Kumar Singh “A survey on 5G: The next generation of mobile communication” Physical Communication, Volume 18, Part 2, March 2016