Jaeseong Jeong

Max-Contribution: On Optimal Resource Allocation in Delay Tolerant Networks

This is by far the first paper considering joint optimization of link scheduling, routing and replication for disruption-tolerant networks (DTNs). The optimization problems for resource allocation in DTNs are typically solved using dynamic programming which requires knowledge of future events such as meeting schedules and durations. This paper defines a new notion of optimality for DTNs, called snapshot optimality where nodes are not clairvoyant, i.e., cannot look ahead into future events, and thus decisions are made using only contemporarily available knowledge. Unfortunately, the optimal solution for snapshot optimality still requires solving an NP-hard problem of maximum weight independent set and a global knowledge of who currently owns a copy and what their delivery probabilities are. This paper presents a new efficient approximation algorithm, called Distributed Max-Contribution (DMC) that performs greedy scheduling, routing and replication based only on locally and contemporarily available information. Through a simulation study based on real GPS traces tracking over 4000 taxies for about 30 days in a large city, DMC outperforms existing heuristically engineered resource allocation algorithms for DTNs.

Delay-Optimal Data Forwarding in Vehicular Sensor Networks

The vehicular sensor network (VSN) is emerging as a new solution for monitoring urban environments such as intelligent transportation systems and air pollution. One of the crucial factors that determine the service quality of urban monitoring applications is the delivery delay of sensing data packets in the VSN. In this paper, we study the problem of routing data packets with minimum delay in the VSN by exploiting 1) vehicle traffic statistics, 2) anycast routing, and 3) knowledge of future trajectories of vehicles such as busses. We first introduce a novel road network graph model that incorporates the three factors into the routing metric. We then characterize the packet delay on each edge as a function of the vehicle density, speed, and the length of the edge. Based on the network model and delay function, we formulate the packet routing problem as a Markov decision process (MDP) and develop an optimal routing policy by solving the MDP. Evaluations using real vehicle traces in a city show that our routing policy significantly improves the delay performance compared with existing routing protocols. Specifically, optimal VSN data forwarding (OVDF) yields, on average, 96% better delivery ratio and 72% less delivery delay than existing algorithms in some areas distant from destinations.

Energy-Efficient Wi-Fi Sensing Policy Under Generalized Mobility Patterns With Aging

An essential condition precedent to the success of mobile applications based on Wi-Fi (e.g., iCloud) is an energy-efficient Wi-Fi sensing. Clearly, a good Wi-Fi sensing policy should factor in both inter-access point (AP) arrival time (IAT) and contact duration time (CDT) distributions of each individual. However, prior work focuses on limited cases of those two distributions (e.g., exponential) or proposes heuristic approaches such as Additive Increase (AI). In this paper, we first formulate a generalized functional optimization problem on Wi-Fi sensing under general inter-AP and contact duration distributions and investigate how each individual should sense Wi-Fi APs to strike a good balance between energy efficiency and performance, which is in turn intricately linked with users mobility patterns. We then derive a generic optimal condition that sheds insights into the aging property, underpinning energy-aware Wi-Fi sensing polices. In harnessing our analytical findings and the implications thereof, we develop a new sensing algorithm, called Wi-Fi Sensing with AGing (WiSAG), and demonstrate that WiSAG outperforms the existing sensing algorithms up to 37% through extensive trace-driven simulations for which real mobility traces gathered from hundreds of smartphones is used.

Max Contribution: An Online Approximation of Optimal Resource Allocation in Delay Tolerant Networks

In this paper, a joint optimization of link scheduling, routing and replication for delay-tolerant networks (DTNs) has been studied. The optimization problems for resource allocation in DTNs are typically solved using dynamic programming which requires knowledge of future events such as meeting schedules and durations. This paper defines a new notion of approximation to the optimality for DTNs, called snapshot approximation where nodes are not clairvoyant, i.e., not looking ahead into future events, and thus decisions are made using only contemporarily available knowledges. Unfortunately, the snapshot approximation still requires solving an NP-hard problem of maximum weighted independent set (MWIS) and a global knowledge of who currently owns a copy and what their delivery probabilities are. This paper proposes an algorithm, Max-Contribution (MC) that approximates MWIS problem with a greedy method and its distributed online approximation algorithm, Distributed Max-Contribution (DMC) that performs scheduling, routing and replication based only on locally and contemporarily available information. Through extensive simulations based on real GPS traces tracking over 4,000 taxies and 500 taxies for about 30 days and 25 days in two different large cities, DMC is verified to perform closely to MC and outperform existing heuristically engineered resource allocation algorithms for DTNs.

TravelMiner: On the Benefit of Path-Based Mobility Prediction

Mobility predictions are becoming more valuable in various applications with the rise of mobile devices. Given that existing prediction techniques are composed of two key procedures: 1) profiling past mobility trajectories as sequences of discrete atomic states (e.g., grid locations, semantic locations) and capturing them with an appropriate statistical model, 2) making a prediction on the next state using the statistical model, TravelMiner tackles the former with paths utilized as the atomic states for the first time, where the paths are defined as sub-trajectories with no branches. Comparing to available location-based predictors, TravelMiner makes a fundamental difference in that it is able to predict the sequence of paths rather than locations, which is far more detailed in the perspective of knowing the exact route to follow. TravelMiner enables this benefit by extracting disjoint paths from GPS trajectories via a similarity metric for curves, called Frechet distance and keeping the sequences of such paths in a statistical model, called probabilistic radix tree. Our extensive simulations over the GPS trajectories of 124 users reveal that TravelMiner outperforms other predictors in diverse popular performance metrics including predictability, prediction accuracy and prediction resolution.

Multi-flow rate control in delayed Wi-Fi offloading systems

Explosive growth of mobile data traffic becomes an increasingly serious problem in cellular networks. Delayed Wi-Fi offloading is the concept to shift the delay-tolerant mobile traffic from cellular networks to Wi-Fi networks at the cost of additional delay. Existing studies mainly focused on a single-flow management or a multi-flow case without specified deadlines. In this paper, we address a multi-flow offloading problem in which a mobile user has multiple traffic flows whose loads and deadlines are different. More precisely, we formulate a multi-flow rate control based on a discrete and finite-horizon Markov decision problem. We develop a dynamic programming (DP)-based optimal rate control algorithm to maximize user satisfaction defined as offloading efficiency minus disutility due to deadline violations. Moreover, we propose a threshold-based rate control algorithm which requires low-complexity and low-memory but achieves high performance. Trace driven simulations based on measurements show that our proposed algorithms achieve high user satisfaction compared to existing algorithms.

Impact of surrounding information on Wi-Fi sensing efficiency

One of essential parts of mobile applications using Wi-Fi is an energy-efficient Wi-Fi sensing. Lately, there have been many studies on Wi-Fi sensing algorithms using surrounding informations (e.g., bluetooth ID [1], cell ID [2] and speed [3], etc). Based on the correlation between such informations and Wi-Fi encounters, the algorithms can determine the wake-up time from sleep mode to detect a Wi-Fi AP efficiently. In this paper, we measure and compare the impact of each surrounding information on the Wi-Fi sensing efficiency by analyzing the uncertainty of information-conditioned remaining time of a mobile until its next Wi-Fi encounter. Using three metrics (i.e., information gain, conditional expectation and variance) on a real mobility trace, we measure to what extend each surrounding information can reduce the uncertainty of the remaining time, which turns out to be the improvement of Wi-Fi sensing efficiency. Comparing such gains from all surrounding informations, we show that the cell ID information is more energy-efficient than bluetooth ID, speed and even all possible combinations of them in Wi-Fi sensing.

DTN Routing Method using Spatial Regularity in Urban Area

The Delay/Disruption Tolerant Network (DTN) is a network designed to operate effectively using the mobility and storage of intermediate nodes under no end-to-end guaranteed network. This new network paradigm is well-suited for networks which have unstable path and long latencies (e.g. interplanetary network, vehicular network). In this paper, we first found that each taxi has its own regularly visiting area and define this property as spatial regularity. We analyze 4000 taxi trace data in Shanghai and show the existence of spatial regularity experimentally. Based on a spatial regularity in urban environment, we present a new DTN routing method. We introduce a Weighted Center (WC) which represents spatial regularity of each node. Through the association with evenly distributed access points (APs) in urban environment, most of vehicles get their grid locations and calculate their WCs. Since our routing method only uses neighbors` WCs for building routing paths, it can be regarded as distributed and practical protocols. Our experiments involving realistic network scenarios created by the traces of about 1500 Shanghai taxies show that our routing method achieves the higher performance compared to ECT, LET by 10%~110%.

Opportunistic shortest path forwarding in delay tolerant networks

Delay Tolerant Networks (DTNs) are characterized by probabilistic links formed among mobile nodes indicating their probabilistic encounters. Prior work on DTN routing uses expected delays as a routing metric to decide the next hop relay node for packet delivery to the destination. However, they measure the expected delays by taking the minimum of the expected delays over all possible paths from a candidate relay. This metric, denoted by MinEx, does not account for the opportunity gain enabled by having multiple paths to the destination through encountering multiple future neighbors. Since DTN routing uses as the relay the first encountered node satisfying given routing criteria, the random delays to multiple relay nodes should be aggregated. Thus, the true expected delays can be measured by taking the expectation of the minimum delays, denoted as ExMin, over all possible probabilistic paths from the candidate.