Memory disaggregation has replaced the landscape of dat-acenters by physically separating compute and memory nodes, achieving improved utilization. As early efforts, kernel paging-based approaches offer transparent virtual memory abstraction for remote memory with paging schemes but suffer from expensive page fault handling. This paper revisits the paging-based approaches and challenges their performance in paging schemes. We posit that the overhead of the paging-based approaches is not a fundamental limitation. We propose DiLOS, a new library operating system (LibOS) specialized for paging-based memory disaggregation. We have revamped the page fault handler to get away with the swap cache and incorporated known techniques in our prefetcher, page manager, and communication module for performance optimization. Furthermore, we provide APIs to augment the LibOS with application semantics. We present two app-aware guides, app-aware prefetching and bandwidth-reducing memory allocator in DiLOS. Through extensive evaluation of microbenchmarks and applications, we demonstrate that DiLOS outperforms the state-of-the-art kernel paging-based system (Fastswap) up to 2.24× and a recent user-level system (AIFM) 1.54× on a real-world data analytic workload.

While mobile instant messaging (MIM) facilitates ubiquitous interpersonal communication, its constant connectivity could build the expectation of an immediate response to messages, and its notifications flood could cause interruptions at inopportune moments. We examine two design concepts for MIM-private status sharing and sender-controlled notifications-that aim to lower the pressure for an immediate reply and reduce unnecessary interruptions by untimely notifications. Private status sharing reactively reveals a customized status with a selected partner(s) only when the partner has sent a message. Sender-controlled notifications give senders the control of choosing whether to send a notification for their own messages. We built MyButler, an Android app prototype that instantiates these two concepts and integrated it with KakaoTalk, a commercial MIM app. During a two-week field study with 11 pairs (5 couples and 6 friend pairs), participants expressed themselves through a total of 210 different statuses, 64.3% of which indicated the current activity or task of the user. Participants reported that private status sharing enabled them to explain their unavailability and relieved the pressure and expectations for timely attendance. We reveal more findings on the types of privately shared statuses and their roles in MIM communication; the in-situ behaviors and patterns of using sender-controlled notifications; and the motivations of MIM users in choosing whether to alert their messages. In terms of message notifications, senders chose to send 25.4% of the messages without any notification. We found that senders' decisions to alert are affected by the receiver's status, their own status to chat, and the possibility of message content exposure to others through notifications. Based on our findings, we draw insights into how the concepts of private status sharing and sender-controlled notifications can be applied in future designs and explorations.

We propose the design of MyButler, a sender-controlled notification management system that mitigates disruption caused by mobile instant messaging through sharing the receiver's activity information with the sender.

In this paper, a relay selection issue is studied to improve data rate performance in a multi-user environment. Previous opportunistic relay selection protocols only consider channel condition in relay selection. However, the target performance is affected by other system attributes, such as delay and load in relay node (RN). The proposed relay selection scheme, which is based on analytic hierarchy process (AHP), adopts signal-to-noise ratio (SNR), switching, and the number of connections to RN as a selection criteria. Also, a special form of eigenvector is proposed and proved analytically. From the proposed eigenvector, the computational complexity is obviously reduced. Target system performance is examined under different weighting parameter values to the above attributes. This work shows that average data rate can be maximized corresponding to the number of RN. Simulation results reveal the benefits of the proposed scheme over the conventional opportunistic relay selection scheme and the threshold-based relay switching selection scheme. The performance of our scheme is verified in terms of average data rate and distribution of the data rate.

In this paper, we address the issue of resource management for coordinated multi-point transmission/reception (CoMP) systems. We propose a dynamic transmission mode selection scheme combined with proportional fair (PF) scheduling, where each cell performs scheduling in a consecutive order. Based on the prior user selection information, each cell decides transmission mode, and makes user scheduling according to the selected transmission mode. Furthermore, we extend precoding matrix indicator (PMI) coordination in order to mitigate intra-cluster interference. The proposed scheme provides adequate resource allocation in JT-CoMP systems without central scheduling unit. Simulation results show that the proposed algorithm provides significant edge user throughput gain without any loss in average throughput.

To encourage data offloading in heterogeneous networks, range expansion techniques have been introduced in LTE-Advanced system. By adding a positive bias in cell selection process, more users can select picocells as their serving cells even the picocell is not of the strongest signal. Also, to protect offloaded users from severe macrocell interference, a specific subframe called almost blank subframe (ABS) is introduced. Since macrocell is kept silent during ABSs, the offloaded users can be allocated without strong macrocell interference within ABSs. It is noted that the offloaded user's average data rate highly depends on the amount of ABSs rather than the bias value because the scheduling opportunities of offloaded users are limited within ABSs. And from the user's point of view, the cell which is expected to give high data rate should be selected as a serving cell. In this paper, we propose a cell selection scheme based on expected user data rate which is a function of the ABS ratio. Simulation results show that the performance of our proposed cell selection scheme can improve average user rate and increase the number of offloaded users as the ABS ratio increases.

In this paper, we deal with a relay selection issue to improve throughput performance under signaling overhead and delay constrained system. The proposed relay selection protocol, which is based on analytic hierarchy process (AHP), adopts signal-to- noise ratio (SNR), system delay, and switching weight as selection criteria. We examined target system performance under various weighting to the above attributes. Also, average throughput can be maximized adaptively in various system environments. The proposed protocol entirely realizes the feature of opportunistic relaying (OR) protocol by totally weighting the SNR criterion. The performance of our protocol is verified in terms of average throughput, average delay and average relay switching counts.

Dual-carrier High Speed Packet Access (DC-HSPA) system has been developed as an alternative method to double data rate using more bandwidth, without use of MIMO. Since it utilizes two carriers at a time, scheduling is more important feature than a single-carrier HSPA (SC-HSPA) system. From the nature of the DC-HSPA, both carriers can be allocated to one user when the user has the highest priority on both carriers. However, if a single carrier is enough to empty the queue, one of carriers has to be selected to transmit own data, while the other carrier is assigned to the other user. Available resource on each carrier varies accordingly to the selected carrier, therefore, efficient carrier selection schemes for DC-HSPA system are proposed based on relative scheduling metric difference between users on each carrier. Simulation results show that proper carrier selection increases throughput as well as average user rate in DC-HSDPA system.

With the growing need of high rate data services, 3GPP standard has investigated ways to support multimedia services with multi-carrier transmissions. Supporting high data rate guaranteeing the quality of service (QoS) and fairness of users is an important issue and should be addressed in an advanced wireless system. This paper investigates packet scheduling considering QoS of real-time traffic and fairness of non-real-time traffic through fixed delay adjustment based on multi-user PF scheduling in a dual-carrier (DC) HSDPA system. The proposed algorithm results in the QoS improvement of real-time traffic and fairness of non-real-time traffic, and the decrease of average sector throughput due to tradeoff between throughput and fairness.

This paper investigates the effect of non-orthogonal decode-and-forward (NDF) relaying scheme in wireless network coding environments. It is well known that network coding is effective to improve overall network performance, but because of error propagation, it may not be always advantageous under a hostile wireless channel. To mitigate error propagation, we adopt NDF relaying scheme in wireless network coding, which allows the source to keep on transmitting the same signal when the relay cooperates to forward the network-coded signal. By NDF relay scheme, SNR of the link between source and destination nodes can be increased, which results in reduced number of errors through the link. Simulation results show that network coding with NDF relay scheme average packet error rate performance as well as goodput in wireless butterfly network.

This paper investigates the effect of combined network coding and power control based on network utility under wireless butterfly networks. Although it is well known that network coding improves network performances, it might not be always advantageous for a hostile wireless environment due to error propagation problem through relaying. Therefore, in this paper, the crossover point for profit-and-loss of network coding and non-network coding according to wireless link quality is derived, and then we show that network utility in terms of spectral efficiency per power is improved, and average power consumption is reduced by adaptively applying relay mode control and power control, which is of bang-bang type with a threshold in wireless butterfly networks.

For maximizing the benefit of network coding in two-way relay networks, the data rate of links from sources to a relay should be the same, since only the same number of received bits can be network-coded at a given time and the rate for broadcast at the relay is limited by the minimum rate over all links. Therefore, we investigate rate control scheme based on power control to maintain the link to be symmetrical while maximizing sum of data rates. We show the impact of rate asymmetry and formulate such problem as a max-min rate optimization problem. A simple iterative solution based on binary search algorithm is provided. Numerical results show the radio resource utilization in terms of bit/sec/Hz/watt is maximized by the proposed rate control scheme when the network coding is applied.

In multicarrier systems, the feedback information from a mobile station to a base station can be used to efficiently allocate the radio resource such as power and/or subchannels to a user. The quality of subchannel is measured by a signal-to-noise ratio (SNR), whose range, in a real system, must be quantized because channel information should be represented as a finite number of bits due to signaling overhead. Therefore, some subchannels whose data rates are the same can exist even though the measured SNRs are different. In this paper, we propose an efficient and fair algorithm to allow users and subchannels to compete through bidding. The bidding price of each user is determined by the difference between the data rate of the best subchannel and the second best, where the highest bidder wins the use of the best subchannel. Simulation results show that the proposed algorithm has better performance than a random selection method in terms of throughput and fairness.

Recently, Wireless Local Area Network (WLAN) mesh networks are attracting significant attention. Because the Institute of Electrical and Electronics Engineers (IEEE) 802.11 Medium Access Control (MAC) protocol has been designed for single- hop networks, the performance of WLAN mesh networks degrades due to congestion, and buffer overflow in intermediate Mesh Nodes (MNs). In this article, we propose new congestion control scheme which applies Adaptive Modulation and Coding (AMC) scheme and effective buffer usage to enhance the performance. AMC scheme uses multiple fragmentation thresholds, modulation levels and coding rates for various data sizes according to Channel State Information (CSI). Because CSI can be used more effectively to squeeze more bits into the medium, the throughput is enhanced. Simulation results show that the proposed scheme achieves much higher throughput than other congestion control schemes.

Vertical handoff is a switching process for seamless communication between heterogeneous wireless networks. By applying pre-handoff concept for the heterogeneous traffic, we propose a two-step vertical handoff algorithm, which can select the best suitable system to a user. From this point of view, we can get a better performance than existing vertical handoff algorithms. Simulation results show that the proposed algorithm increases the throughput and decreases the handoff rate and the connection time ratio with target systems.

In this paper, we propose a packet-by-packet scheduling algorithm which grants priority to each packet instead of a user. The proposed scheduler classifies packets inside of each user into three classes such as emergency, near emergency, and non-emergency packets based on packet delay ratio, and applies different scheduling schemes to grant priority in each region respectively. It can be possible to manage packets more efficiently for guaranteeing various Quality of Services (QoS) in wireless multimedia systems. Simulation results show that the proposed scheme offers higher system throughput and lower packet drop rate than other scheduling algorithms.