Swarup Mandal

A technique to support dynamic pricing strategy for differentiated cellular mobile services

With the advent of a multi-operator regime in the sector of cellular mobile services, users became sensitive to the quality of service (QoS), and, as a result service providers started offering differentiated services. Each operator shall ensure the promised QoS to every subscriber even during busy hours of a day when the network operates at the maximum load. During off-peak hours the load on the network decreases. As a result, the network designed to meet the QoS requirement during busy hours of a day leaves a large amount idle resource during off-peak hours of the day. In this scenario, a service provider faces the problem of maximizing revenue while satisfying resource and QoS constraints. A dynamic differentiated pricing strategy (DDPS) for the cellular mobile service is a solution to the above problem. In this paper, we have proposed a DDPS which, to the best of our knowledge, is the first of its kind in the literature. We have compared the performance of our proposed solution with the static differentiated pricing strategies (SDPSs) with respect to revenue earned and average network resource utilization. The experimental results show that our proposed solution provides a substantial improvement over the SDPSs

Pricing wireless network services using smart market models

The growing popularity of wireless network has prompted the service providers to look for appropriate ways to meet the demand while recovering the invested capital. This calls for a good revenue model, where pricing of services plays an important role. This paper investigates the performance of smart market based pricing models for wireless network services. Conventionally, these models have been successfully applied in various other service fields including wired networks. In particular, we apply the French auction and Dutch auction models and compare their performances in comparison with the flat-pricing scheme, which is quite popular in the wireless domain. Simulation results on the performance the proposed schemes, with respect to the total revenue generated and the call blocking probability, show encouraging results.

Heuristic search techniques for cell to switch assignment in location area planning for cellular networks

The issue of grouping cells into location areas (LAs), where each LA is serviced by a switch, plays an important role in the planning of cellular networks. It is a combinatorial optimization problem that is known to be NP-hard. This paper proposes a total cost of operation (TCO) minimizing state space search formulation of the problem and a heuristic for assigning cell to switches. The TCO includes both recurring hand-off cost and amortized fixed cost. The proposed heuristic is used with block depth first search (BDFS), and also with iterative deepening A (IDA) to solve the problem. Detailed experiments show that the BDFS has a better performance than IDA with respect to execution time while finding an optimal solution. BDFS also outperforms other existing techniques that are based on meta-heuristics, namely, simulated annealing (SA), genetic algorithm (GA), tabu search (TS), and H-I, in terms of solution quality when they are constrained to run for a user specified time.

A real-time heuristic search technique for fixed channel allocation (FCA) in mobile cellular communications

Abstract Fixed Channel Allocation (FCA) is an important issue in cellular communications due to its direct impact on an optimal usage of scarce radio resources and also on user-perceived QoS. In this paper, we have formulated the FCA as a minimum cost set cover problem and proposed a lower bound heuristic to solve it by a heuristic search technique, namely, the block depth-first search (BDFS). The BDFS with a lower bound heuristic is capable of finding: (a) a satisfactory solution under real-time constraints, and (b) an optimal solution when time constraint is relaxed. Experiments reveal that the BDFS produces considerably improved results, when compared to meta-heuristic techniques, such as the Genetic Algorithm, the Simulated Annealing, and the Tabu Search.

Post-deployment tuning of UMTS cellular networks through dual-homing of RNCs

In conventional UMTS cellular networks, during deployment usually a set of NodeBs is assigned to one Radio Network Controller (RNC), and a set of RNCs to one Serving GPRS Support Node (SGSN) for data services, as well as to one Mobile Switching Centre (MSC) for voice services. Operators thus far have considered single-homing of RNCs to MSCs/SGSNs (i.e., many-to-one mapping) with an objective to reduce the total cost over a fixed period of time. However, a single-homing network does not remain cost-effective any more when subscribers later on begin to show specific inter-MSC/SGSN mobility patterns (say, diurnality of office goers) over time. This necessitates post-deployment topological extension of the network in terms of dual-homing of RNCs, in which some specific RNCs are connected to two MSCs/SGSNs via direct links resulting in a more complex many-to-two mapping structure in parts of the network. The partial dual-homing attempts to increase link cost minimally and reduce handoff cost maximally, thereby significantly reducing the total cost in a post-deployment optimal extension. In this paper, we formulate the scenario as a combinatorial optimization problem and solve the NP-Complete problem using two meta-heuristic techniques, namely Simulated Annealing (SA) and Tabu search (TS). We then compare these techniques with a novel optimal heuristic search method that we propose typically to solve the problem. The comparative results reveal that, though all of them perform equally well for small networks, for larger networks, the search-based method is more efficient than meta-heuristic techniques in finding optimal solutions quickly.

A novel direction-based diurnal mobility model for handoff estimation in cellular networks

In real-life scenario where users mostly know their destinations a priori, their movement is not purely random (as assumed in random walk model); rather it is somewhat direction-based and depends on the present location and the final destination. For instance, an office goer often returns home in the evening (i.e., mobility is diurnal). While no existing research has studied direction-based diurnal mobility model which could be of immense importance to cellular operators, we formulate a Markov model for such movement and give a theoretical upper bound of cell boundary crossings (i.e., expected number of handoffs) by the user. We have theoretically calculated the number of handoffs based on our proposed mobility model and verified the result in a simulated environment. We find that a majority of inter-RNC/MSC handoffs [4] comes from only a few cells which should be made dual-homed by the operators for their cost reduction.

Dynamic Price Discovering Models for Differentiated Wireless Services

Heterogeneous subscriber base with wide range of wireless services makes the wireless service market more challenging for a service provider (SP). This challenge is related to the revenue model of the SP, where pricing policy plays a central role. The pricing policy should be such that an SP will recover the investment while keeping the desired level of customer satisfaction. An ideal pricing policy should be less complex and be able to price the service close to the worth, a subscriber attached to it. This calls for dynamic price discovering models for differentiated wireless service. In this paper, we have proposed auction based price discovering models like uniform pricing auction and discriminatory pricing auction for dynamic pricing of differentiated wireless services. We have compared their performances with flat-pricing scheme for differentiated wireless services, which is quite popular in the wireless domain. Simulation results show that auction based models have a potential to replace the traditional flat pricing models.

Optimization of probability of false alarm and probability of detection in cognitive radio networks using GA

In this paper, we optimize probability of detection and probability of false alarm in cognitive radio network to minimize probability of error of a particular SU in a centralized cognitive radio network using Genetic algorithm (GA). Our objective is to minimize probability of error and find out optimum values of probability of occupancy detection or probability of detection and probability of false alarm. We use Genetic Algorithm to solve this optimization problem. The result is compared with Differential Evolution algorithm and it is evident from the comparison that DE finds better solution and takes much lesser number of evaluations to find optimum solution.

A sustainable and collaborative strategy for dynamic spectrum management in next generation wireless networks

Next generation wireless technologies offer various services from voice call to full motion pictures and even to high speed internet access. Consequently, the service providers (SP) armed with different wireless technologies (like 2.5G/3G/LTE) would require an adequate and significant amount of spectrum bandwidth for satisfying the need of their customers. Hence to achieve complete commercialization, the SPs, operating simultaneously, would demand for more and more spectrum from the regulatory body of the country. The spectrum demand on the part of the SP may vary with time (dynamic) because of varied kind of loads which are generated depending on the nature of the client-base, their requirements and their expected quality of experience. This work has addressed this challenging issue of allocating spectrum dynamically to different technologies under the portfolio of an SP. Here, we have conceived a scenario where service providers (SP) own multiple access networks (ANs) of different technologies. We envisage that an entity, called local spectrum controller (LSC) which is dedicated for managing the common pool of spectrum allocated to each SP. LSC is mainly responsible for distributing the spectrum to individual ANs of an SP in a fair manner. Since the available spectrum may not be sufficient enough to satisfy the aggregate demand from all ANs simultaneously, an LSC may face a situation, where satisfying individual demands from all ANs may result in a compromise between the demand and supply. This demand-supply situation would force an LSC or an SP to adhere to some dynamic spectrum management strategy, where demands of an AN would have to be satisfied depending on the current state of available spectrum and required usage of it. This calls for an adaptive dynamic strategy to be introduced by an SP for efficient spectrum distribution. The dynamic disparity of spectrum allocation can be idealized as a game between LSC and ANs. Hence, in the present work, we have modeled the problem of dynamic spectrum allocation as an n-player cooperative bankruptcy game and have solved the problem with the help of Shapley value and @t-value separately. We have investigated whether the ANs find it beneficial to cooperate with each other to make the solution sustainable enough. To evaluate the performances of the games that the ANs play, we have designed a novel utility function for each AN. We have identified plausible aims of an SP as minimizing overall dissatisfaction (MOD) and maximizing equality of distribution (MED). Next, we have studied performances of the above two solution concepts against max-min fairness algorithm (benchmarked in our case) with respect to the above objectives of LSC. Finally, we have proposed a unique heuristic in order to facilitate the decision making process of dynamic spectrum allocation, which leads to an adaptive yet optimized spectrum allocation strategy.

Joint optimization of 2-tier dual-homing for NodeBs and RNCs in UMTS networks using meta-heuristic techniques

Conventional design of UMTS networks usually involves single-homing (i.e., many-to-one mapping) of NodeBs to Radio Network Controllers (RNCs) in tier-1 (i.e., a group of NodeBs is connected to a single RNC) and that of RNCs to Mobile Switching Centres (MSCs) as well as to Serving GPRS Support Nodes (SGSNs) in tier-2 (i.e., a group of RNCs is connected to a single MSC/SGSN). Thus, any NodeB is connected to only one RNC and any RNC is connected to only one MSC/SGSN. However, as subscriber distribution changes over time and new mobility patterns of subscribers begin to evolve, single homing solution sometimes becomes inefficient in terms of handoff cost minimization. One solution to this brown-field operational problem is dual-homing extension of some selected NodeBs and RNCs (i.e., some NodeBs are connected to two RNCs in tier-1 and some RNCs to two MSCs/SGSNs in tier-2) in order to reduce the handoff cost. Traditionally, this optimization problem has been formulated separately for each tier and solved independently, thereby missing the global optimal solution. In this paper, we have first shown how to combine the optimization problems across the two tiers and then mapped the joint dual homing optimization problem into a classical search problem. Next, we have used two common meta-heuristic techniques, namely Simulated Annealing and Tabu Search, to solve the above problem. Comparison of the results obtained from joint dual homing with the published results for individual dual homing reveals that the joint dual homing performs considerably better than individual dual homing that attacks NodeB level and RNC level separately and independently.