Abstract With the popularization of cloud computing, various cloud services have emerged, and hybrid clouds that can take advantage of combining public and private clouds are attracting attention. However, because of their variety, determining a combination of cloud services suited to the user’s current environment and requirements is expensive when deploying a hybrid cloud. Even if the required services are available, there is a lack of tools to connect them, manage them in batches, and utilize the integrated environment. To solve these problems, this paper proposes a cloud selection and integration process (C-SIP), which selects and integrates a combination of cloud services through a hybrid cloud service broker (hybrid CSB), which is an automation solution supporting hybrid cloud deployment. Moreover, the proposed method is realized using a script including the application programming interface of each cloud service. The proposed C-SIP will be used as a core approach toward the hybrid CSB, which is expected to facilitate the introduction of hybrid clouds and the acquisition of cloud strategies.

We consider the Markov Decision Process (MDP) of selecting a subset of items at each step, termed the Select-MDP (S-MDP). The large state and action spaces of S-MDPs make them intractable to solve with typical reinforcement learning (RL) algorithms especially when the number of items is huge. In this paper, we present a deep RL algorithm to solve this issue by adopting the following key ideas. First, we convert the original S-MDP into an Iterative Select-MDP (IS-MDP), which is equivalent to the S-MDP in terms of optimal actions. IS-MDP decomposes a joint action of selecting K items simultaneously into K iterative selections resulting in the decrease of actions at the expense of an exponential increase of states. Second, we overcome this state space explosion by exploiting a special symmetry in IS-MDPs with novel weight shared Q-networks, which provably maintain sufficient expressive power. Various experiments demonstrate that our approach works well even when the item space is large and that it scales to environments with item spaces different from those used in training.

Rate Adaptation (RA) is a fundamental mechanism in 802.11 systems. It allows transmitters to adapt the coding and modulation scheme as well as the MIMO transmission mode to the radio channel conditions, to learn and track the (mode, rate) pair providing the highest throughput. The design of RA mechanisms has been mainly driven by heuristics. In contrast, we rigorously formulate RA as an online stochastic optimization problem. We solve this problem and present G-ORS (Graphical Optimal Rate Sampling), a family of provably optimal (mode, rate) pair adaptation algorithms. Our main result is that G-ORS outperforms state-of-the-art algorithms such as MiRA and Minstrel HT, as demonstrated by experiments on a 802.11n network test-bed. The design of G-ORS is supported by a theoretical analysis, where we study its performance in stationary radio environments where the successful packet transmission probabilities at the various (mode, rate) pairs do not vary over time, and in non-stationary environments where these probabilities evolve. We show that under G-ORS, the throughput loss due to the need to explore sub-optimal (mode, rate) pairs does not depend on the number of available pairs. This is a crucial advantage as evolving 802.11 standards offer an increasingly large number of (mode, rate) pairs. We illustrate the superiority of G-ORS over state-of-the-art algorithms, using both trace-driven simulations and test-bed experiments.

The data explosion and resource scarcity of mobile cellular networks require new paradigms to effectively integrate heterogeneous radio resources. Of many candidate approaches, smart aggregation of LTE and Wi-Fi radios is a promising solution that bonds heterogeneous links to meet a mobile terminal's bandwidth need. Motivated by the existence of a significant number of carrier operated Wi-Fi APs, we propose an easily deployable mechanism, called LTE-W, which efficiently utilizes LTE and Wi-Fi links only with the minimum change of eNodeBs, LTE backhaul networks, and mobile terminals. LTE-W, which is a link-level aggregation mechanism, has the following two key components: 1) mode selection and 2) bearer-split scheduling. First, in the mode selection, LTE-W internally decides who should be served by either LTE-only or LTE-Wi-Fi aggregation considering intra-cell fairness rather than just following users' intention of aggregation. For the users' preference to be offered the aggregation service, we choose a bearer (roughly defined in LTE as a set of flows with a similar QoS) as a basic unit of aggregation and propose a smart intra-bearer scheduling algorithm that splits a bearer's traffic into LTE and Wi-Fi links, considering the performance of TCP flows that take two heterogeneous wireless links. We evaluate our mechanism using the NS-3 with LENA, under various configurations, including nodes with mobility and HTTP traffic, and compare it with a transport-level aggregation mechanism, multipath TCP (MPTCP), demonstrating that LTE-W significantly improves MPTCP, e.g., up to 75% in terms of Jain's fairness index.

People are seeking solutions in diverse directions to cope with mobile data explosion and resource scarcity in mobile cellular networks. Of many candidate approaches, smart aggregation of LTE and Wi-Fi radios is a promising solution that bonds heterogeneous links to meet a mobile terminal's available bandwidth need. Motivated by the existence of a significant number of carrier-operated Wi-Fi APs, we propose a mechanism, called LTE-W, of efficiently utilizing LTE and Wi-Fi links only with the minimum changes of eNodeBs, LTE backhaul networks, and mobile terminals. Our mechanism has the following two key components: (i) mode selection and (ii) bearer-split scheduling. In the mode selection, LTE-W internally decides who should be served by either of LTE or LTE-Wi-Fi aggregation considering intra-cell fairness rather than just following users' intention of aggregation. For the users decided to be offered the aggregation service, we choose a bearer (roughly defined a set of flows with a similar QoS in LTE) as a basic unit of aggregation and propose a smart intra-bearer scheduling algorithm that splits a bearer's traffic into LTE and Wi-Fi links, considering the tuning of TCP flows that take two heterogeneous wireless links. We evaluate our mechanism using the NS-3 with LENA, and compare it to a transport-level aggregation mechanism, MPTCP, demonstrating that LTE-W significantly improves MPTCP, e.g., up to 75% in terms of Jain's fairness index.

In the past few years, an exciting progress has been made on CSMA (Carrier Sense Multiple Access) algorithms that achieve throughput and utility optimality for wireless networks. However, most of these algorithms are known to exhibit poor delay performance making them impractical for implementation. Recently, several papers have addressed the delay issue of CSMA and yet, most of them are limited, in the sense that they focus merely on specific network scenarios with certain conditions rather than general network topology, achieve low delay at the cost of throughput reduction, or lack rigorous provable guarantees. In this paper, we focus on the recent idea of exploiting multiple channels (actually or virtually) for delay reduction in CSMA, and prove that it is per-link delay order-optimal, i.e., O(1)-asymptotic-delay per link, if the number of virtual channels is logarithmic with respect to mixing time of the underlying CSMA Markov chain. The logarithmic number is typically small, i.e., at most linear with respect to the network size. In other words, our contribution provides not only a provable framework for the multiple-channel based CSMA, but also the required explicit number of virtual-multi-channels, which is of great importance for actual implementation. The key step of our analytic framework lies in using quadratic Lyapunov functions in conjunction with (recursively applying) Lindley equation and Azuma's inequality for obtaining an exponential decaying property in certain queueing dynamics. We believe that our technique is of broader interest in analyzing the delay performance of queueing systems with multiple periodic schedulers.

Rate Adaptation (RA) is a fundamental mechanism in 802.11 systems. It allows transmitters to adapt the coding and modulation scheme as well as the MIMO transmission mode to the radio channel conditions, and in turn, to learn and track the (mode, rate) pair providing the highest throughput. So far, the design of RA mechanisms has been mainly driven by heuristics. In contrast, in this paper, we rigorously formulate such design as an online stochastic optimisation problem. We solve this problem and present ORS (Optimal Rate Sampling), a family of (mode, rate) pair adaptation algorithms that provably learn as fast as it is possible the best pair for transmission. We study the performance of ORS algorithms in stationary radio environments where the successful packet transmission probabilities at the various (mode, rate) pairs do not vary over time, and in non-stationary environments where these probabilities evolve. We show that under ORS algorithms, the throughput loss due to the need to explore sub-optimal (mode, rate) pairs does not depend on the number of available pairs. This is a crucial advantage as evolving 802.11 standards offer an increasingly large number of (mode, rate) pairs. We illustrate the efficiency of ORS algorithms (compared to the state-of-the-art algorithms) using simulations and traces extracted from 802.11 test-beds.

In the past few years, an exciting progress has been made on CSMA (Carrier Sense Multiple Access) algorithms that achieve throughput and utility optimality for wireless networks. However, most of these algorithms are known to exhibit poor delay performance making them impractical for implementation. Recently, several papers have addressed the delay issue of CSMA and yet, most of them are limited, in the sense that they focus merely on specific network scenarios with certain conditions rather than general network topology, achieve low delay at the cost of throughput reduction, or lack rigorous provable guarantees. In this paper, we focus on the recent idea of exploiting multiple channels (actually or virtually) for delay reduction in CSMA, and prove that it isper-link delay orderoptimal, i.e.,O(1)-asymptotic-delay per link, if the number of virtual channels is logarithmic with respect to mixing time of the underlying CSMA Markov chain. The logarithmic number is typically small, i.e., at most linear with respect to the network size. In other words, our contribution provides not only a provable framework for the multiple-channel based CSMA, but also the required explicit number of virtual-multi-channels, which is of great importance for actual implementation. The key step of our analytic framework lies in using quadratic Lyapunov functions in conjunction with (recursively applying) Lindley equation and Azuma's inequality for obtaining an exponential decaying property in certain queueing dynamics. We believe that our technique is of broad interest in analyzing the delay performances of other general queueing systems.