Andreas Pamboris

Optimized query routing trees for wireless sensor networks

In order to process continuous queries over Wireless Sensor Networks (WSNs), sensors are typically organized in a Query Routing Tree (denoted as T) that provides each sensor with a path over which query results can be transmitted to the querying node. We found that current methods deployed in predominant data acquisition systems construct T in a sub-optimal manner which leads to significant waste of energy. In particular, since T is constructed in an ad hoc manner there is no guarantee that a given query workload will be distributed equally among all sensors. That leads to data collisions which represent a major source of energy waste. Additionally, current methods only provide a topological-based method, rather than a query-based method, to define the interval during which a sensing device should enable its transceiver in order to collect the query results from its children. We found that this imposes an order of magnitude increase in energy consumption. In this paper we present MicroPulse^+, a novel framework for minimizing the consumption of energy during data acquisition in WSNs. MicroPulse^+ continuously optimizes the operation of T by eliminating data transmission and data reception inefficiencies using a collection of in-network algorithms. In particular, MicroPulse^+ introduces: (i) the Workload-Aware Routing Tree (WART) algorithm, which is established on profiling recent data acquisition activity and on identifying the bottlenecks using an in-network execution of the critical path method; and (ii) the Energy-driven Tree Construction (ETC) algorithm, which balances the workload among nodes and minimizes data collisions. We show through micro-benchmarks on the CC2420 radio chip and trace-driven experimentation with real datasets from Intel Research and UC-Berkeley that MicroPulse^+ provides significant energy reductions under a variety of conditions thus prolonging the longevity of a wireless sensor network.

ETC: Energy-Driven Tree Construction in Wireless Sensor Networks

Continuous queries in Wireless Sensor Networks (WSNs) are founded on the premise of Query Routing Tree structures (denoted as T), which provide sensors with a path to the querying node. Predominant data acquisition systems for WSNs construct such structures in an ad-hoc manner and therefore there is no guarantee that a given query workload will be distributed equally among all sensors. That leads to data collisions which represent a major source of energy waste. In this paper we present the Energy-driven Tree Construction (ETC) algorithm, which balances the workload among nodes and minimizes data collisions, thus reducing energy consumption, during data acquisition in WSNs. We show through real micro-benchmarks on the CC2420 radio chip and trace-driven experimentation with real datasets from Intel Research and UC-Berkeley that ETC can provide significant energy reductions under a variety of conditions prolonging the longevity of a wireless sensor network.

Towards enabling hyper-responsive mobile apps through network edge assistance

Poor Internet performance currently undermines the efficiency of hyper-responsive mobile apps such as augmented reality clients and online games, which require low-latency access to real-time backend services. While edge-assisted execution, i.e. moving entire services to the edge of an access network, helps eliminate part of the communication overhead involved, this does not scale to the number of users that share an edge infrastructure. This is due to a mismatch between the scarce availability of resources in access networks and the aggregate demand for computational power from client applications. Instead, this paper proposes a hybrid edge-assisted deployment model in which only part of a service executes on LTE edge servers. We provide insights about the conditions that must hold for such a model to be effective by investigating in simulation different deployment and application scenarios. In particular, we show that using LTE edge servers with modest capabilities, performance can improve significantly as long as at most 50% of client requests are processed at the edge. Moreover, we argue that edge servers should be installed at the core of a mobile network, rather than the mobile base station: the difference in performance is negligible, whereas the latter choice entails high deployment costs. Finally, we verify that, for the proposed model, the impact of user mobility on TCP performance is low.

AD-APT: blurring the boundary between mobile advertising and user satisfaction

The choice between two dominant monetization strategies for mobile apps, i.e. capitalising on revenue generated from either users or advertisers, is not a straightforward one. While users are initially attracted more to free (ad-supported) apps, paid (ad-free) apps are more likely to persist over time by avoiding the adverse implications of mobile advertising on device energy and network usage. This paper proposes a system that strikes a balance between the two extremes: AD-APT re-factors ad-supported apps automatically to adjust the frequency of mobile ad occurrences at runtime based on policies that consider the device’s battery life, the type of network connectivity and limits on network usage. We evaluate AD-APT on ten popular ad-supported Android apps and show that it can yield reductions of up to 30× in network usage caused by mobile ads and 40% in energy consumption.

Demo:: NOMAD: An Edge Cloud Platform for Hyper-Responsive Mobile Apps

Fast access to backend services is crucial for many mobile apps. For example, emerging augmented-reality devices such as Google Glass require fast access to powerful servers to achieve seamless interactivity with the real world; and online gaming clients need to communicate in real-time through centralised game services. A major obstacle to achieving this hyper responsiveness is the performance of the underlying network that interconnects mobile clients and services. Network e↵ects cannot be anticipated, let alone controlled, due to the unpredictability of wide-area networks and the fact that users roam between di↵erent networks. Over time, organisations have gone to great lengths to reduce access latency to backend services by moving them “closer” to end users. In 2010, Google spent

Better Performance in LTE Networks with Edge Assistance: The World of Warcraft Case

1.9 billion on a data centre in New York, despite real estate prices being amongst the highest in the world, to gain direct access to local and global networks [5]. Similarly, cloud service providers such as Amazon AWS have rolled out new infrastructure in edge locations. Proactive measures against high network latencies, however, are limited by the fact that network proximity in a mobile setting is unknown a priori. To enable hyper-responsive mobile apps, Balan et al. [1] first proposed cyber-foraging, i.e. the use of remote resources to augment smartphone capabilities. This led to proposals such as Cloudlets [7], which treat smartphones as thin clients served by virtual device clones, and systems such as MAUI [4] and CloneCloud [2], which apply a more finegrained app partitioning to reduce response times. In general, the above approaches can only improve application responsiveness to the extent that computation delays dominate performance. It remains a challenge to control the impact of high network latencies, especially when users roam between wireless networks of di↵erent operators.

PortLand: a scalable fault-tolerant layer 2 data center network fabric

To improve the performance of Massively Multiplayer Online Games (MMOGs) in mobile networks, we explore the potential benefits of an edge-assisted deployment model: part of the MMOG backend service executes closer to the end user at the edge of the LTE network. We investigate the impact on game latency of (1) the exact placement of such edge servers; (2) the number of cooperating game clients; (3) the amount of client requests served at the network edge; (4) the hardware capabilities of edge servers; and (5) user roaming. Based on our analysis, we show that edge assistance can in fact increase the performance of online games over LTE networks as long as at most 50% of the user requests are processed at the network edge. Furthermore, we argue that the Packet Data Network Gateway (PGW) is the most appropriate place for hosting edge servers and show that TCP performance in the proposed setting is not affected by user roaming.