To be compatible with the legacy 802.11, there are two major medium access control (MAC) behaviors, high throughput (HT) and non-high throughput (non-HT), in the 802.11n. In this paper, we analyze and compare the energy efficiencies of different MAC behaviors in 802.11n on the basis of the Bianchi model and our previous works to evaluate the performance of the different MAC behaviors regarding HT and non-HT. Our studies try to provide the decision for the mobile stations to enable the HT of 802.11n or not based on the consideration of energy efficiency. Studies show that owing to the large power consumption in HT, it is not suitable for limited power devices to carry WWW traffics by multiple-input multiple-output transmission because of large overheads of physical layer in the HT mode. However, if large file transmissions by File Transfer Protocol are considered, the energy efficiency in HT MAC can be very high because of the large aggregated frame size. It is especially true when the number of active stations is large because of the decrease in idle listening time by using the techniques applied in HT MAC such as Aggregate MAC Protocol Data Unit and Block-ACK. These characteristics in the HT mode can overwhelm the larger overheads of physical layer compared with that in the non-HT mode when large files are needed to be uploaded. Copyright © 2012 John Wiley & Sons, Ltd.

Topology construction, the initial step of the topology control, is an important technique for a wireless ad hoc network to be energy-efficient. In [4], we have proposed to use the relative neighborhood graph (RNG) to obtain an RNG-based topology in which the transmission ranges between wireless nodes are reduced. Then, among the RNG-based topology, we proposed a green clustering algorithm (GCA) to organize the wireless nodes into a clustered network topology. In this paper, we further analyze the energy consumption in exchanging data packets and cluster maintenance messages. Simulation results confirm that the proposed RGCA (i.e., combining the RNG and GCA) provides a way to construct an energy-efficient cluster topology for wireless ad hoc networks.

LTE has been touted as a leading-edge mobile communication technology offering high data rate and low latency. However, with a sophisticated physical layer to boost performance, the processing demand on user equipment (UE) is tremendous, which implies that the energy consumption is heavy. To curb undesirable energy waste and extend battery life, a variety of energy-conserving measures have been developed, among which discontinuous reception (DRX) is shown to be useful. This paper introduces the light sleeping mode in order to further improve the performance of DRX for traffic with QoS requirements in LTE networks. The key idea of light sleeping is to turn off the power amplifier but leave the other components in the transceiver on to cut down energy consumption while allowing fast wakeup. Quantitative analysis shows that the proposed method can substantially reduce energy consumption and satisfy the delay requirement of QoS traffic.

802.11 Wireless Local Area Network (WLAN) technology is now common in power sensitive devices like smart phones and personal digital assistants (PDAs). In this article, we present an accurate power consumption model based on the Bianchi model [13] and the power measurement in PHY layer [9] to predict the power consumption of 802.11n and Multiple-Inputs-Multiple-Outputs (MIMO) mode of 802.11n. In this model, we calculate the total power consumption by summing six consumed powers: the idle listening power consumed in the DIFS period, the back-off stage, the power consumed in transmitting a frame, the idle listening power consumed in the SIFS turn-around time, the power consumed in receiving an ACK frame from AP and the power consumed in the collision for one frame transmission. Our analyses and simulations show that the power consumed in idle listing will dominate the total power consumption when the number of the active stations is large. The successful possibility and MAC efficiency of 802.11n are also analyzed and simulated in this article. The imperfect channel scenario will be evaluated, too. This power analysis model to our knowledge has not been presented in the previous works so far.

Energy consumption issue is an important research topic in wireless ad hoc networks, because wireless nodes in such networks operate on limited battery power. This paper describes two novel routing algorithms that take into account the effect of distance between nodes and network congestion level on energy consumption. Our algorithms adapt existing AODV routing protocol to improve performance in terms of energy conservation and other performance metrics. We formulate the routing problem as the minimization of energy cost functions which reflect the key features of the problem under consideration. Simulation results indicate that the routing schemes are more efficient than existing well-known routing protocols.

In order to execute a parallel PDE (partial differential equation) solver on a shared-memory multiprocessor, we have to avoid memory conflicts in accessing multidimensional data grids. A new multicoloring technique is proposed for speeding sparse matrix operations. The new technique enables parallel access of grid-structured data elements in the shared memory without causing conflicts. The coloring scheme is formulated as an algebraic mapping which can be easily implemented with low overhead on commercial multiprocessors. The proposed multicoloring scheme bas been tested on an Alliant FX/80 multiprocessor for solving 2D and 3D problems using the CGNR method. Compared to the results reported by Saad (1989) on an identical Alliant system, our results show a factor of 30 times higher performance in Mflops. Multicoloring transforms sparse matrices into ones with a diagonal diagonal block (DDB) structure, enabling parallel LU decomposition in solving PDE problems. The multicoloring technique can also be extended to solve other scientific problems characterized by sparse matrices.

A new multicoloring technique is proposed for parallel sparse matrix-vector multiplication, which dominates the computing cost of iterative PDE (partial differential equation) solvers. The new technique enables parallel solution of grid-structured nonsymmetric PDE problems on shared-memory multiprocessors through resolving memory access conflicts by multiple processors. The coloring scheme is formulated as an algebraic mapping which can be implemented with low overhead.