Demand proportional resource provisioning schemes have been proposed to achieve datacenter energy efficiency, where servers are turned on/off according to the load of requests. Most existing schemes focus on delay sensitive jobs (SENs) only. However, in datacenters, there exist a vast amount of delay-tolerant jobs (TOLs), such as background/maintenance jobs. Thus, we study joint SEN and TOL resource provisioning in this paper, with a focus on TOLs. We consider traffic dynamics of SENs and TOLs in different time scales, and electricity price temporal dynamics and location diversity. Our goal is to minimize total costs, while guaranteeing QoS for SENs and achieving a desirable delay performance for TOLs. Specifically, we propose a joint server provisioning, SEN load dispatching, TOL load shifting, and SEN/TOL capacity allocation scheme, which leverages TOL queue information and does not assume any system statistical information. We also design other benchmark schemes that leverage different system information. Both analytical results and extensive simulation results show the efficiency of the proposed scheme, named OrgQ, in reducing total costs and TOL queue delay.

In data centers, traffic demand varies in both large and small time scales. A data center with dynamic traffic often needs to over-provision active servers to meet the peak demand, which incurs significant energy cost. In this paper, our goal is to reduce energy cost of a set of distributed Internet data centers (IDCs) while maintaining the quality of service of the dynamic traffic. In particular, we consider the outage probability as the QoS metric, where outage is defined as service demand exceeding the capacity. We require the outage probability at each IDC to be smaller than a predefined threshold. Our goal is thus to minimize total energy cost over all IDCs, subject to the outage probability constraint. We achieve the goal by dynamically adjusting server capacity and performing load shifting in different time scales. We propose three different load-shifting and joint capacity allocation schemes with different complexity and performance. Our schemes leverage both stochastic multiplexing gain and electricity-price diversity. Thus, improving over prior work, our schemes reduce energy consumption/cost even when all IDCs have the same electricity price. We use both simulated load traces and real traffic traces to evaluate the performance of the proposed schemes. Results show that our proposed schemes are efficient in reducing energy cost, and robust in QoS provisioning.

Market mechanisms have been exploited as important means for spectrum acquisition and access in cognitive radio networks. In this paper, we propose a two-tier market for decentralized dynamic spectrum access. In the proposed Tier-1 market, spectrum is traded from a primary user (PU) to secondary users (SUs) in a relatively large time scale to reduce signaling overhead. Then, driven by dynamic traffic demands, SUs set up the Tier-2 market to redistribute channels among themselves in a small time scale. More specifically, we use a Nash bargain game to model the spectrum acquisition of SUs in the Tier-1 market and derive the equilibrium prices. We then employ a strategic bargain game to study the spectrum redistribution in the Tier-2 market, where SUs can exchange channels with low overhead through random matching, bilateral bargain, and the predetermined market equilibrium price. We investigate how various factors, such as the availability of channels and bargain partners, matching strategies, and traffic dynamics, affect the market relationships. This work provides new understanding on the spectrum market and valuable guidelines to primary and secondary network operators.

To reduce datacenter energy consumption and cost, current practice has considered demand-proportional resource provisioning schemes, where servers are turned on/off according to the load of requests. Most existing work considers instantaneous (Internet) requests only, which are explicitly or implicitly assumed to be delay-sensitive. On the other hand, in datacenters, there exist a vast amount of delay-tolerant jobs, such as background/maintainance jobs. In this paper, we explicitly differentiate delay-sensitive jobs and delay tolerant jobs. We focus on the problem of using delay-tolerant jobs to fill the extra capacity of datacenters, referred to as trough/valley filling. Giving a higher priority to delay-sensitive jobs, our scheme complements most existing demand-proportional resource provisioning schemes. Our goal is to design an intelligent trough filling mechanism that is energy efficient and also achieves good delay performance. Specifically, we propose a joint dynamic speed scaling and traffic shifting scheme. The scheme does not need statistical information of the system, which is usually difficult to obtain. In the proposed scheme, energy cost saving comes from dynamic speed scaling, statistical multiplexing, electricity price diversity, and service efficiency diversity. In addition, good delay performance is achieved via load shifting and capacity allocation based on queue conditions. We show the efficiency of the proposed scheme by both analysis and simulation.

Cognitive radio (CR) improves spectrum efficiency by allowing secondary users (SUs) to dynamically exploit the idle spectrum owned by primary users (PUs). This paper studies optimal bandwidth allocation of SUs for throughput efficiency. Consider the following tradeoff: an SU increases its instantaneous throughput by accessing more spectrum, but channel access/switching overhead, contention among multiple SUs, and dynamic PU activity create higher liability for larger bandwidths. So how much is too much? In this paper, we study the optimal bandwidth allocation for multiple SUs. Our approach is twofold. We first study the optimal bandwidth an SU should use to maximize the per-SU throughput in the long term. The optimal bandwidth is derived in the context of dynamic PU activity, where we consider both independent and correlated PU channel scenarios while accounting for the effects of channel switching overhead. We further consider the case of suboptimal spectrum use by SUs in the short term due to PU activity dynamics. We propose an efficient channel reconfiguration (CREC) scheme to improve SUs' performance. We use real PU channel activity traces in the simulations to validate our results. The work sheds light on the design of spectrum sharing protocols in cognitive radio networks.

In this paper, our goal is to achieve an optimal tradeoff between energy efficiency and service performance over a set of distributed IDCs with dynamic demand. In particular, we consider the outage probability as the QoS metric, where outage is defined as service demand exceeding the capacity of an IDC. Our goal is thus to minimize total energy cost over all IDCs, subject to the outage probability constraint. We achieve the goal by dynamically adjusting server capacity and performing load shifting in different time scales. We propose three different load-shifting and joint capacity allocation schemes with different complexity and performance. Our schemes leverage both stochastic multiplexing gain and electricity-price diversity.

Market mechanisms have been exploited as important means for spectrum acquisition and access in cognitive radio networks. In this paper, we propose a two-tier market for decentralized dynamic spectrum access. In the proposed Tier-1 market, spectrum is traded from a primary user (PU) to secondary users (SUs) in a relatively large time scale to reduce signaling overhead. Then driven by dynamic traffic demands, SUs set up the Tier-2 market to redistribute channels among themselves in a small time scale. More specifically, we use Nash bargain game to model the spectrum acquisition of SUs in the Tier-1 market and derive the equilibrium prices. We then use strategic bargain game to study the spectrum redistribution in the Tier-2 market, where SUs can exchange channels with low overhead through random matching, bilateral bargain, and the predetermined market equilibrium prices. We disclose how various factors, such as availability of channels and bargain partners, matching schemes, and traffic dynamics, impact the market relationships. This work provides better understanding on the spectrum market and valuable guidelines to primary and secondary network operators.

Previous results show that a node's throughput scales poorly as the network size increases when every node has traffic. However, in many cases, only a fraction of nodes in large networks have data to send or receive at any given time, while other nodes can act as relays/routers. Therefore, in this paper, we study the scaling behavior from a user's viewpoint (a user is a node with traffic). We first derive an upper bound on per user throughput. To derive the lower bound, we propose a simple scheduling scheme that enables users to cooperate with relay nodes and fully utilize the networks capacity. We show that per user throughput depends on the network size, the number of users, and the node deployment schemes, and is in general much more optimistic. Our scheme also sheds light on designing efficient cooperation protocols in heterogeneous networks and cognitive radio networks.

In this paper, we propose a new and rather simple grouped-subcarrier allocation algorithms with proportional fairness among users in downlink OFDM transmission. The proposed algorithm tries to minimize the required transmit power while satisfying the rate requirement and BER constraint of each user. Subcarrier and power allocation are performed in two steps. We are mainly focusing on the subcarrier allocation step. First subcarriers are grouped to reduce the computational complexity of the allocation algorithm. Then we employ proportional constraints in terms of rate ratios to assure each user to achieve the target data rate in advance. Simulation results show that this low complexity resource allocation scheme achieves better performance than the fixed frequency division approach and shows comparable performance compared to previously derived suboptimal resource distribution schemes. It is also shown that with rate constraints the capacity is distributed more fairly and rationally among users