Item exchange is becoming a popular behavior and widely supported in more and more online community systems, e.g. online games and social network web sites. Traditional manual search for possible exchange pairs is neither efficient nor effective. Automatic exchange pairing is increasingly demanding in such community systems, and potentially leading to new business opportunities. To meet the needs on item exchange in the market, each user in the system is entitled to list some items he/she no longer needs, as well as some required items he/she is seeking for. Given the values of all items, an exchange between two users is eligible if 1) they both have some unneeded items the other one wants, and 2) the exchange items from both sides are approximately of the same total value. To efficiently support exchange recommendation services, especially with frequent updates on the listed items, new data structures are proposed in this paper to maintain promising exchange pairs for each user. Extensive experiments on both synthetic and real data sets are conducted to evaluate our proposed solutions.

Traditional bipolar differential amplifiers have only a ±5mV operational range with nonlinear distortion below 0.1dB. In this paper, a linearization technique based on neural network training algorithm is proposed to expand this 0.1dB linear region to a much wider ±200mV range. Compared with the traditional and recent state-of-the-art techniques for linearization, where gain or noise performance is always being tradeoff for high linearity, the proposed technique leads to the increase of the amplifier's gain with low noise. Another advantage of the proposed approach is that, it is implemented using a simple, paralleled structure that leads to a smaller area and less power consumption than other linearization schemes. The design procedure also enables a completely customizable solution based on required linear range, ripples, and the number of available transistors. The experimental results show that the proposed linearized amplifier has a very good linearity and is capable to wide bandwidth, good gain, and low noise performance.

No abstract

Abstract. There is an increasing interest in developing Phase Change Memory (PCM) based main memory systems. In order to retain the latency benefits of DRAM, such systems typically have a small DRAM buffer as a part of the main memory. However, for these systems to be widely adopted, limitations of PCM such as low write endurance and expensive writes need to be addressed. In this paper, we propose PCMaware sorting algorithms that can mitigate writes on PCM by efficient use of the small DRAM buffer. Our performance evaluation shows that the proposed schemes can significantly outperform existing schemes that are oblivious of the PCM.

This paper implements a "multi-tanh principle" to linearize the bipolar differential amplifiers by offsetting the transfer function of the differential pairs and connection them in parallel. The principle is analyzed in detail and the optimum 3.75 and 2.9 offset ratio is found for a double and a 4-cell linearized differential amplifier. The resulted 4-cell linearized amplifier achieves an extremely linear 20mV dynamic range within less than 0.1 dB ripple. The structure is shouldable for further expansion for wider linear operation. Temperature dependence is discussed and compensation from 250K-400K is realized by implementing the transistor size ratio, the Widlar current source and MOS voltage shift cell. Such differential amplifier with high linearity and insensitivity to stress will be suitable for high frequency situation or accurate gain control in varies of industrial applications.