Studies On Call Admission Control And Handover Techniques For High Altitude Platform Communications Systems

In High Altitude Platform Communications System (HAPCS), communications payload is put on the High Altitude Platform (HAP) which operates at altitudes between 17 to 22 km in the stratosphere. Since the operating altitude of the HAP is between that of the terrestrial radio tower and the satellite, HAPCS possesses the advantages of both terrestrial wireless communications systems and satellite communications systems. Generally, HAP utilizes multiple-beam antenna to project cells on the ground, and improves the system capacity through frequency reusing among different cells. The position instability and the replacement of the HAP inevitably bring on the inter-cell handover and the inter-HAP handover, affect the system performance greatly, and become a bottleneck which restricts the application of the HAPCS. This thesis discusses the Call Admission Control (CAC) and handover problems in the HAPCS. CAC and handover schemes are designed to reduce/mitigate the handover failures introduced by the platform movements, and to reduce the call blocking probability. Meanwhile, good handover dropping performance can also benefit to relaxing the requirements of the communications payload to the platform instability and mobility.Platform movements, such as horizontal/vertical displacement and yaw, force the calls in the service area handover between adjancent cells. This thesis establishes the equations to calculate the maximum horizontal/vertical HAP displacement, which accurately describe the relationship between the maximum horizontal/vertical displacement and system prameters, such as user antenna beamwidth and platform altitude. Also, we establish the equations to calculate the average call handover probability introduced by three kinds of platform movements mentioned in the above section. These equations describe the relationships between the average call handover probability and system parameters, such as maximum HAP displacement or maximum HAP deviation angle. In addition, a HAP will return to the ground for maintenance, where the platform replacement is inevitable. Especially, the platform replacement is frequent for a short endurance HAP based HAPCS. This thesis establishes the inter-HAP handover model introduced by the platform replacement. Based on the cell overlap condition and the user beamwidth condition, the equations are established to calculate inter-HAP handover performance, such as handoff success ratio, Handover Opportunity Interval (HOI), HOI start rate and handover load. We illustrate the relationship between the inter-HAP handover performance and system parameters, such as user antenna beamwidth, platform altitude and user handover scheme.In a cellular system, the traffic in adjacents cells is non-uniform, which means that when one cell is blocked, its adjacent cells may probabily carry light traffic. Traffic transfer based CAC and handover schemes reduce the call blocking probability and/or the handover dropping probability through transferring the traffic in the blocked cell to the adjacent un-blocked cell. Moreover, the handover introduced by the platform movement in a HAPCS has a characteristic of directionality, which means that the handover direction can be determined by the moving characteristic of the HAP when only fixed users exist in the service area. Exploiting the traffic non-uniformity and the directional handover characteristics, this thesis proposes a Cooperative Directional Handover Scheme (CDHS) based on traffic transfer and handover queue schemes. CDHS tries to execute a call transfer operation before the queue time of a handover request runs out, and to avoid handover failure due to queue time out of the handover request in the cell which this transferred call resides in. Comparing with handover queue scheme, this scheme reduces the handover dropping probability greatly. In addition, we propose a directional traffic aware handover cheme based on the load balancing scheme (LBS) and the CDHS, which executes the load balancing operation when the traffic difference between the current cell and one of its adjacent cells exceeds a specific threshold, and utilizes the CDHS in the handover process. Comparing with LBS and CDHS, this scheme can achieve a better unified system performance when the platform speed is high and the power factor of the call blocking probability is moderate, and can achieve good tradeoff between call blocking and handover dropping performance.Presently, guaranteed handover scheme can guarantee the services with high priority to avoid handover failures, but it introduces a high call blocking probability. This thesis proposes an overlap area assisted guaranteed handover scheme, which exploits both the geographical information and the overlap area information to judge and avoid the handover failures introduced by the new incoming call in the call admission process. For one or two types of services, this scheme can improve the call blocking performance significantly while maintains a zero handover dropping probability seen with the Time based Channel Reservation Algorithm (TCRA). Furthermore, for the HAPCS adopting an adaptive modulation and coding scheme, we propose a rate transition area assisted guaranteed handover scheme, which introduces a Rate Transition Area (RTA) between higher rate region and lower rate region. The new call blocking performance can be improved through making the calls with lower rate in the RTA handover earlier to the higher rate, or making the calls with higher rate in the RTA handover later to the lower rate. For the services which allow a specific handover dropping probability, a handover scheme can achieve good call blocking and handover dropping performance through making the best of the geographical information and considering the possibility of call leaving. We propose a handover time interval restricted CAC scheme according to the traffic distribution, which restricts the handover dropping probability by limiting the N+1th handover time interval in a virtual cell if the number of the calls in that virtual cell exceeds cell capacity N. Comparing with dynamic doppler-based handover prioritization scheme, this scheme can achieve better call blocking performance while a smaller handover dropping probability threshold is chosen. A handover performance restricted CAC is proposed which exploits the traffic distribution information to estimate the handover dropping probability introduced by a new incoming call, and restricts this handover dropping probability with a specific threshold. This scheme can maintain the handover dropping probability to a specific threshold even for a high traffic load condition.Frequency sharing can achieve high spectrum efficiency when a HAPCS and a terrestrial cellular system coexist. However, in a two-layer cellular system, frequency sharing can introduce strong cross-layer co-channel interference between co-channel macrocells and microcells, and can even result in unsuccessful communication for the users in part of the coverage area. This thesis proposes the partial coverage area frequency sharing scheme, in which the areas with strong cross-layer co-channel interference, namely Cross-layer Interference Limited Areas (CILAs), can be intelligently determined with the assistance of the cognitive users adopting directional antennas, and separate frequency bands are allocated to the CILAs of the macrocells and the microcells. By properly selecting the sizes of the CILAs in the macrocells and the microcells, this scheme can achieve good spectrum efficiency and an acceptable link outage probability, even if the users adopt low directional antennas. Hence this cheme can eliminate the unsuccessful communication problem of the users in part of the coverage area for the whole coverage area frequency sharing scheme.Double coverage in the two-layer cellular system can be regarded as an extention to the cell overlap in the single-layer cellular system. The call blocking performance of the guaranteed handover scheme can be improved greatly through properly exploiting this double coverage characteristic. In a multiple cellular system, overflow is an effective way to balance the traffic between different cell layers, and take-back can eliminate the problems caused by the mismatch of the cell size and the traffic category. This paper extends the TCRA to the two-layer cellular system by utilizing the overflow operation, tack-back operation and a set of traffic flows comprised of several ordered overflow operations. The geographical information, double coverage characteristic, call overflow and/or call take-back between adjacent cell layers are exploited to judge and to avoid the handover failures introduced by the new incoming call in the call admission process. Overflow operation can help improving call blocking performance greatly, and the more the overflow operations are exploited, the better the call blocking performance. Tack-back operation can further help improving the call blocking performance, and greatly reducing the overflowed traffic.