Achla Marathe

Characterizing the interaction between routing and MAC protocols in ad-hoc networks

We empirically study the effect of mobility and interaction between various input parameters on the performance of protocols designed for wireless ad-hoc networks. An important objective is to study the interaction of the routing and MAC layer protocols under different mobility parameters. We use three basic mobility models: grid mobility model, random waypoint model, and exponential correlated random model. The performance of protocols is measured in terms of various quality of service measures including (i) latency, (ii) throughput, (iii) number of packets received and (iv) long term fairness. Three different commonly studied routing protocols are used: AODV, DSR and LAR scheme 1. Similarly three well known MAC protocols are used: MACA, 802.11 and CSMA.Our main contribution is simulation based experiments coupled with emph rigorous statistical analysis to characterize the emph interaction between the above stated parameters. Such methods allow us to analyze complicated experiments with large input space in a systematic manner. From our results, we conclude the following: No single MAC or routing protocol dominated the other protocols in their class. More interestingly, no MAC/routing protocol combination was better than other combinations over all mobility models and response variables.In general, it is not meaningful to speak about a MAC or a routing protocol in isolation. Presence of interaction leads to trade-offs between the amount of control packets generated by each layer. The results raise the possibility of improving the performance of a particular MAC layer protocol by using a cleverly designed routing protocol or vice-versa..

'Beating the news' with EMBERS: forecasting civil unrest using open source indicators

We describe the design, implementation, and evaluation of EMBERS, an automated, 24x7 continuous system for forecasting civil unrest across 10 countries of Latin America using open source indicators such as tweets, news sources, blogs, economic indicators, and other data sources. Unlike retrospective studies, EMBERS has been making forecasts into the future since Nov 2012 which have been (and continue to be) evaluated by an independent T&E team (MITRE). Of note, EMBERS has successfully forecast the June 2013 protests in Brazil and Feb 2014 violent protests in Venezuela. We outline the system architecture of EMBERS, individual models that leverage specific data sources, and a fusion and suppression engine that supports trading off specific evaluation criteria. EMBERS also provides an audit trail interface that enables the investigation of why specific predictions were made along with the data utilized for forecasting. Through numerous evaluations, we demonstrate the superiority of EMBERS over baserate methods and its capability to forecast significant societal happenings.

Liquidity and stock returns in emerging equity markets

Abstract Using data for 27 emerging equity markets for the period January 1992 through December 1999, we document the behavior of liquidity in emerging markets. We find that stock returns in emerging countries are positively correlated with aggregate market liquidity as measured by turnover ratio, trading value and the turnover–volatility multiple. The results hold in both cross-sectional and time-series analyses, and are quite robust even after we control for world market beta, market capitalization and price-to-book ratio. The positive correlation between stock returns and market liquidity in a time-series analysis is consistent with the findings in developed markets. However, the positive correlation in a cross-sectional analysis appears to be at odds with market microstructure theory that has been empirically supported by studies on developed markets. Our findings regarding the cross-sectional relation between stock returns and liquidity is consistent with the view that emerging equity markets have a lower degree of integration with the global economy.

Cascading failures in multiple infrastructures: From transportation to communication network

This research conducts a systematic study of human-initiated cascading failures in critical inter-dependent societal infrastructures. The focus is on three closely coupled systems: (i) cellular and mesh networks, (ii) transportation networks and (iii) social phone call networks. We analyze cascades that occur in inter-dependent infrastructures due to behavioral adaptations in response to a crisis. During crises, changes in individual behavioral lead to altered calling patterns and activities, which influence the urban transport network. This, in turn, affects the loads on wireless networks. The interaction between these systems and their co-evolution poses significant technical challenges for representing and reasoning about these systems. We develop interaction-based models in which individuals and infrastructure elements are placed in a common geographic coordinate system. The goal is to study the impact of a chemical plume in a densely populated urban region. Authorities order evacuation of the affected area which leads to change in peoples activity patterns as they are forced to drive home or to evacuation shelters. They also use the wireless networks for coordination among family members and information sharing. These two behavioral adaptations, cause flash-congestion in the urban transport network and the wireless network. We analyze how extended periods of unanticipated road congestion can result in failure of infrastructures, starting with the servicing base stations in the congested area. Finally, we study the criticality and robustness of the various base stations and measure how congestion in the transportation network impacts communication infrastructure.

An Integrated Modeling Environment to Study the Co-evolution of Networks, Individual Behavior and Epidemics

We discuss an interaction-based approach to study the coevolution between socio-technical networks, individual behaviors, and contagion processes on these networks. We use epidemics in human population as an example of this phenomenon. The methods consist of developing synthetic yet realistic national-scale networks using a first principles approach. Unlike simple random graph techniques, these methods combine real world data sources with behavioral and social theories to synthesize detailed social contact (proximity) networks. Individual-based models of within-host disease progression and inter-host transmission are then used to model the contagion process. Finally, models of individual behaviors are composed with disease progression models to develop a realistic representation of the complex system in which individual behaviors and the social network adapt to the contagion. These methods are embodied within Simdemics – a general purpose modeling environment to support pandemic planning and response. Simdemics is designed specifically to be scalable to networks with 300 million agents – the underlying algorithms and methods in Simdemics are all high-performance computing oriented methods. New advances in network science, machine learning, high performance computing, data mining and behavioral modeling were necessary to develop Simdemics. Simdemics is combined with two other environments, Simfrastructure and Didactic, to form an integrated cyberenvironment. The integrated cyber-environment provides the end-user flexible and seamless Internet based access to Simdemics. Service-oriented architectures play a critical role in delivering the desired services to the end user. Simdemics, in conjunction with the integrated cyber-environment, has been used in over a dozen user defined case studies. These case studies were done to support specific policy questions that arose in the context of planning the response to pandemics (e.g., H1N1, H5N1) and human initiated bio-terrorism events. These studies played a crucial role in the continual development and improvement of the cyber-environment.

Analyzing interaction between network protocols, topology and traffic in wireless radio networks

We study the interaction between communication protocols, network topology and packet traffic in wireless static radio networks. A particular interest is to empirically characterize the effect of interaction between the routing layer and the MAC layer on overall system performance. Three well-known MAC protocols: 802.11, CSMA and MACA are considered. Similarly three recently proposed routing protocols: AODV, DSR and LAR scheme 1 are considered. The performance of the protocols is measured with regard to three important parameters: (i) number of packets received, (ii) average latency of each packet and (iii) long term fairness. We use a simple statistical technique based on ANOVA (analysis of variance), to characterize the effect of interaction between protocols and various input parameters on network performance. This technique is of independent interest and can be utilized in other simulation studies. Using our methodology, we conclude that different combinations of routing and MAC protocols yield varying performance under varying network topology and traffic situations. Our results show that no combination of routing protocol and MAC protocol is the best over all situations. An important implication of the study is that the performance analysis of protocols at a given level in the protocol stack needs to be studied not locally in isolation but as a part of the complete protocol stack.

Assessing the efficiency of US electricity markets

Abstract The recent California’s energy crisis has raised doubts about the benefits of energy deregulation. While it is true that the California electricity market is in turmoil, other electricity markets like the Pennsylvania–NewJersey–Maryland (PJM) are doing fine. This paper assesses the mark of efficiency reached by the electricity markets in California, New York, and PJM. It also compares the degree of efficiency across markets (forward vs. real time) and across time. No significant differences between the California and PJM electricity markets were discovered in the year of California’s energy crisis (2000) using the cointegration tests. This research suggests that differences in price behavior between these two markets during 2000 did not arise from differences in efficiency. According to our analysis and measures of efficiency, PJM and California electricity markets are more efficient than the New York market. Also, as these markets become more mature over time, their efficiency level goes up. We also found evidence that a multi-settlement scheduling system leads to higher efficiency.

Sensitivity of Household Transmission to Household Contact Structure and Size

Objective Study the influence of household contact structure on the spread of an influenza-like illness. Examine whether changes to in-home care giving arrangements can significantly affect the household transmission counts. Method We simulate two different behaviors for the symptomatic person; either s/he remains at home in contact with everyone else in the household or s/he remains at home in contact with only the primary caregiver in the household. The two different cases are referred to as full mixing and single caregiver, respectively. Results The results show that the household’s cumulative transmission count is lower in case of a single caregiver configuration than in the full mixing case. The household transmissions vary almost linearly with the household size in both single caregiver and full mixing cases. However the difference in household transmissions due to the difference in household structure grows with the household size especially in case of moderate flu. Conclusions These results suggest that details about human behavior and household structure do matter in epidemiological models. The policy of home isolation of the sick has significant effect on the household transmission count depending upon the household size.

Forecasting Social Unrest Using Activity Cascades

Social unrest is endemic in many societies, and recent news has drawn attention to happenings in Latin America, the Middle East, and Eastern Europe. Civilian populations mobilize, sometimes spontaneously and sometimes in an organized manner, to raise awareness of key issues or to demand changes in governing or other organizational structures. It is of key interest to social scientists and policy makers to forecast civil unrest using indicators observed on media such as Twitter, news, and blogs. We present an event forecasting model using a notion of activity cascades in Twitter (proposed by Gonzalez-Bailon et al., 2011) to predict the occurrence of protests in three countries of Latin America: Brazil, Mexico, and Venezuela. The basic assumption is that the emergence of a suitably detected activity cascade is a precursor or a surrogate to a real protest event that will happen “on the ground.” Our model supports the theoretical characterization of large cascades using spectral properties and uses properties of detected cascades to forecast events. Experimental results on many datasets, including the recent June 2013 protests in Brazil, demonstrate the effectiveness of our approach.

Coevolution of epidemics, social networks, and individual behavior: a case study

This research shows how a limited supply of antivirals can be distributed optimally between the hospitals and the market so that the attack rate is minimized and enough revenue is generated to recover the cost of the antivirals. Results using an individual based model find that prevalence elastic demand behavior delays the epidemic and change in the social contact network induced by isolation reduces the peak of the epidemic significantly. A microeconomic analysis methodology combining behavioral economics and agent-based simulation is a major contribution of this work. In this paper we apply this methodology to analyze the fairness of the stockpile distribution, and the response of human behavior to disease prevalence level and its interaction with the market.