Reconstruction-based Model-Based Reinforcement Learning (MBRL) agents, such as Dreamer, often fail to discard task-irrelevant visual distractions that are prevalent in natural scenes. In this paper, we propose a reconstruction-free MBRL agent, called DreamerPro, that can enhance robustness to distractions. Motivated by the recent success of prototypical representations, a non-contrastive self-supervised learning approach in computer vision, DreamerPro combines Dreamer with prototypes. In order for the prototypes to benefit temporal dynamics learning in MBRL, we propose to additionally learn the prototypes from the recurrent states of the world model, thereby distilling temporal structures from past observations and actions into the prototypes. Experiments on the DeepMind Control suite show that DreamerPro achieves better overall performance than state-of-the-art contrastive MBRL agents when there are complex background distractions, and maintains similar performance as Dreamer in standard tasks where contrastive MBRL agents can perform much worse.

In this paper, we propose an efficient multihop broadcast protocol for asynchronous duty-cycled wireless sensor networks (EMBA) where each node independently wakes up according to its own schedule. EMBA adopts two techniques of the forwarder's guidance and the overhearing of broadcast messages and ACKs. A node transmits broadcast messages with guidance to neighbor nodes. The guidance presents how the node forwards the broadcast message to neighbor nodes by using unicast transmissions. This technique significantly reduces redundant transmissions and collisions. The overhearing of broadcast messages and ACKs helps to reduce the number of transmissions, thus it minimizes the active time of nodes. We implement EMBA and conventional protocols of ADB and RIMAC broadcast in ns-2 simulator to compare their performance. The simulation results show that EMBA outperforms ADB and RI-MAC broadcast in both sparse and dense networks. EMBA achieves lower message cost than the conventional protocols and significantly improves the energy efficiency in terms of both duty cycle and energy consumption.

To reduce the number of unnecessary handover in hierarchical macro/femto-cell networks, it is necessary to avoid macro → femto cell handovers of temporary femtocell visitors who stay in the femtocell for a relatively short time. In this paper, we propose a smart handover decision algorithm exploiting future mobility pattern prediction scheme to prevent macro → femto cell handovers of such temporary femtocell visitors. Our simulation result shows that the proposed algorithm effectively reduces the number of unnecessary handovers.

MapReduce is a programming model that supports distributed and parallel processing for large-scale data-intensive applications such as machine learning, data mining, and scientific simulation. Hadoop is an open-source implementation of the MapReduce programming model. Hadoop is used by many companies including Yahoo!, Amazon, and Facebook to perform various data mining on large-scale data sets such as user search logs and visit logs. In these cases, it is very common to share the same computing resources by multiple users due to practical considerations about cost, system utilization, and manageability. However, Hadoop assumes that all cluster nodes are dedicated to a single user, failing to guarantee high performance in the shared MapReduce computation environment. In this paper, we propose two optimization schemes, prefetching and pre-shuffling, which improve the overall performance under the shared environment while retaining compatibility with the native Hadoop. The proposed schemes are implemented in the native Hadoop-0.18.3 as a plug-in component called HPMR (High Performance MapReduce Engine). Our evaluation on the Yahoo!Grid platform with three different workloads and seven types of test sets from Yahoo! shows that HPMR reduces the execution time by up to 73%.