Ashwin Sridharan - Academia.edu (original) (raw)
Papers by Ashwin Sridharan
Lecture Notes in Computer Science, 2007
As mobile devices grow more powerful and diverse in the applications they run, it becomes increas... more As mobile devices grow more powerful and diverse in the applications they run, it becomes increasingly desirable to give them some independence in scheduling their transmissions. The latest release of the CDMA2000 1xEV-DO system incorporates a number of mechanisms that afford mobile devices some flexibility in making uplink transmission decisions. With distributed decisions, however, comes the possibility of harmful interactions that could lower overall system efficiency. In this paper, we carry out a preliminary investigation of a proposed token bucket based mechanism for enabling distributed transmission decisions. We explore a class of simple probabilistic policies that readily lend themselves to distributed operation, and show that their efficacy greatly depends on the number of devices sharing the uplink. Specifically, we identify regions in the number of active devices in which both the optimal (in terms of overall system throughput) transmission policy and the device token efficiency (the achieved transmission rate per token) exhibit significantly different properties. Our findings provide initial insight into how to configure token bucket parameters and transmission policies in order to achieve a reasonable compromise between distributed decisions and overall system efficiency.
2009 17th IEEE International Conference on Network Protocols, 2009
Proceedings of the 14th ACM international conference on Mobile computing and networking - MobiCom '08, 2008
We present an experimental characterization of the physical and MAC layers in CDMA 1xEV-DO and th... more We present an experimental characterization of the physical and MAC layers in CDMA 1xEV-DO and their impact on transport layer performance. The 1xEV-DO network is currently the fastest mobile broadband cellular network, offering data rates of up to 3.1 Mbps for both stationary and mobile users. These rates are achieved by using novel capacity enhancement techniques at the lower layers. Specifically, 1xEV-DO incorporates rapid channel rate adaptation in response to signal conditions, and opportunistic scheduling to exploit channel fluctuations. Although shown to perform well in isolation, there is no comprehensive literature that examines the impact of these features on transport layer and application performance in real networks.
This paper investigates the impact of traffic granularity on the performance of routing algorithm... more This paper investigates the impact of traffic granularity on the performance of routing algorithms that incorporate traffic information. Our focus is on robustness of routing performance. The granularity of a traffic matrix can affect network robustness through two opposing factors. Fine granularity traffic matrices allow flexibility in load balancing over a large number of paths by virtue of a large number of streams. This improves the ability of the network to absorb traffic fluctuations. However, as we shall show in this work, increasing the granularity of matrix can increase the variability of the streams which degrades the robustness of a routing solution. Moreover, increasing the granularity of the matrix may also increase the forwarding state cost due to an increase in the number of classifiers and paths. We investigate the trade-off between these two opposing factors to determine the appropriate operating point in terms of traffic granularity. Our analysis is based on traffi...
Proceeding of IEEE International Symposium on a World of Wireless, Mobile and Multimedia Networks 2014, 2014
Lecture Notes in Computer Science, 2007
Modern CDMA wireless channels support multiple transmission rates, which can be dynamically assig... more Modern CDMA wireless channels support multiple transmission rates, which can be dynamically assigned to users based on traffic demand. However, in practice, assignment of high rate channels comes with the penalty of increased power as well as smaller orthogonal codes, which constrains their assignment to only a subset of active users. This motivates the need to carefully control high rate channel assignments so as to minimize power and achieve fairness among users. In this work, we propose a simple class of channel allocation policies to achieve this goal for TCP sessions. We develop an analytical model that explicitly captures both TCP dynamics and the impact of multiple users contending for a shared resource to evaluate the performance of the allocation policy. The model is shown to be accurate by comparing against ns-2 simulations and its utility demonstrated by computing the minimum number of required high rate channels to minimize contention.
Proceedings of the 12th annual international conference on Mobile computing and networking - MobiCom '06, 2006
TCP is the dominant transport protocol over both wired and wireless links. It is however, well kn... more TCP is the dominant transport protocol over both wired and wireless links. It is however, well known that TCP is not suitable for wireless networks and several solutions have been proposed to rectify this shortcoming. In this work we explore cross-layer optimization of the rate adaptation feature of cellular networks to optimize TCP throughput. Modern cellular networks incorporate RF technology that allows them to dynamically vary the wireless channel rate in response to user demand and channel conditions. However, the set of data rates as well as the scheduler policy are typically chosen to optimize throughput for inelastic applications. In order to optimize such a system for TCP, we propose a two state TCP-aware scheduler that switches between two rates as a function of the TCP sending rate. We develop a fluid model of the steady-state TCP behavior for such a system and derive analytical expressions for TCP throughput that explicitly account for rate variability as well as the dependency between the scheduler and TCP. Using the model we choose RF layer parameters that, in conjunction with the TCP-aware scheduler, improve long-term TCP throughput by the order of 15−25%. We also compare our analytical results against those obtained from ns-2 simulations and confirm that our model indeed closely approximates TCP behavior in such an environment.
Proceedings of the third ACM international workshop on Wireless network testbeds, experimental evaluation and characterization - WiNTECH '08, 2008
We describe our experiences in building a city-wide infrastructure for wide-area wireless experim... more We describe our experiences in building a city-wide infrastructure for wide-area wireless experimentation. Our infrastructure has two components -(i) a vehicular testbed consisting of wireless nodes, each equipped with both cellular (EV-DO) and WiFi interfaces, and mounted on city buses plying in Madison, Wisconsin, and (ii) a software platform to utilize these testbed nodes to continuously monitor and characterize performance of large scale wireless networks, such as city-wide mesh networks, unplanned deployments of WiFi hotspots, and cellular networks. Beyond our initial efforts in building and deploying this infrastructure, we have also utilized it to gain some initial understanding of the diversity of user experience in large-scale wireless networks, especially under various mobility scenarios. Since our vehicle-mounted testbed nodes have fairly deterministic mobility patterns, they provide us with much needed performance data on parameters such as RF coverage and available bandwidth, as well as quantify the impact of mobility on performance. We use our initial measurements from this testbed to showcase its ability to provide an efficient, low-cost, and robust method to monitor our target wireless networks. These initial measurements also highlight the challenges we face as we continue to expand this infrastructure. We discuss what these challenges are and how we intend to address them.
2014 IEEE 22nd International Conference on Network Protocols, 2014
Cellular providers are rapidly deploying multiple technologies like cell biasing, carrier aggrega... more Cellular providers are rapidly deploying multiple technologies like cell biasing, carrier aggregation, co-ordinated interference control/scheduling to improve capacity and coverage. In this paper, we explore a complementary transport layer approach based on multipath TCP that can concurrently use multiple interfaces to boost throughput of users with poor coverage and improve fairness. Multipath TCP has been recently standardized by IETF and requires no modifications to applications. It has been shown to improve fairness and throughput in wireline environments and individual user throughputs in wireless networks. However, in a wireless multi-user environment, it is not clear that it is always beneficial, as we show in this paper. Therefore, we examine if it is indeed beneficial for a service provider to judiciously decide whether to enable multiple cellular interfaces on a smartphone based on a global centralized view of its network. Alternatively, should a device decide independently based only on a local view? To quantify the network wide impact in a system where users have multiple cellular interfaces, we have developed centralized and distributed heuristic algorithms to evaluate this, particularly in the context of fairness across all the users. Our simulations and numerical models show that there are potential gains in fairness (15-30%) to be realized by judiciously enabling multipath connections at the cell edge. These gains diminish as the number of users in a cell increases or users behave greedily. We also quantify the delicate balance between throughput and fairness. Our analysis provides an intuition on which user(s) in a cellular network stand to benefit the most by enabling multiple interfaces. We also discuss LTE protocol mechanisms to enforce associations of specific interfaces to specific cells.
2010 IEEE Wireless Communication and Networking Conference, 2010
We consider a number of related problem formulations pertaining to adaptive subcarrier allocation... more We consider a number of related problem formulations pertaining to adaptive subcarrier allocation in multiuser Orthogonal Frequency-Division Multiplexing (OFDM) systems, and prove that they are NP-hard. Thus there exist no known algorithms that can provide optimal solutions for all instances of these problems in polynomial time. We further prove that these problems are hard to approximate in polynomial time. Finally, we discuss qualitatively the settings under which worstcase performance is likely to be observed. If P = NP , then a problem is as easy to compute as to verify it. Consequently, for all the well-known hard NP-complete problems, such as Traveling Salesman, there would would exist an algorithm which can guarantee optimal solutions in polynomial computation time. It is generally conjectured that P = NP . For more details on NP-hardness, see [10].
Proceeding of the 14th ACM SIGKDD international conference on Knowledge discovery and data mining - KDD 08, 2008
Lecture Notes in Computer Science, 2005
An important requirement of a robust traffic engineering solution is insensitivity to changes, be... more An important requirement of a robust traffic engineering solution is insensitivity to changes, be they in the form of traffic fluctuations or changes in the network topology because of link failures. In this paper we focus on developing a fast and effective technique to compute traffic engineering solutions for Interior Gateway Protocol (IGPs) environments that are robust to link failures in the logical topology. The routing and packet forwarding decisions for IGPs is primarily governed by link weights. Our focus is on computing a single set of link weights for a traffic engineering instance that performs well over all single logical link failures. Such types of failures, although usually not long lasting, of the order of tens of minutes, can occur with high enough frequency, of the order of several a day, to significantly affect network performance. The relatively short duration of such failures coupled with issues of computational complexity and convergence time due to the size of current day networks discourage adaptive reactions to such events. Consequently, it is desirable to a priori compute a routing solution that performs well in all such scenarios. Through computational evaluations we demonstrate that our technique yields link weights that perform well over all single link failures and also scales well, in terms of computational complexity, with the size of the network.
IEEE INFOCOM 2014 - IEEE Conference on Computer Communications, 2014
ABSTRACT Proportional Fair (PF) scheduling algorithms are the de-facto standard in cellular netwo... more ABSTRACT Proportional Fair (PF) scheduling algorithms are the de-facto standard in cellular networks. They exploit the users' channel state diversity (induced by fast-fading), and are optimal for stationary channel state distributions and an infinite time-horizon. However, mobile users experience a non-stationary channel, due to slow-fading (on the order of seconds), and are associated with basestations for short periods. Hence, we develop the Predictive Finite-horizon PF Scheduling ((PF)2S) Framework that exploits mobility. We present extensive channel measurement results from a 3G network and characterize mobility-induced channel state trends. We show that a user's channel state is highly reproducible and leverage that to develop a data rate prediction mechanism. We then present a few channel allocation estimation algorithms that rely on the prediction mechanism. Our trace-based simulations consider instances of the PF2S Framework composed of combinations of prediction and channel allocation estimation algorithms. They indicate that the framework can increase the throughput by 15%-55% compared to traditional PF schedulers, while improving fairness.
Lecture Notes in Computer Science, 2007
As mobile devices grow more powerful and diverse in the applications they run, it becomes increas... more As mobile devices grow more powerful and diverse in the applications they run, it becomes increasingly desirable to give them some independence in scheduling their transmissions. The latest release of the CDMA2000 1xEV-DO system incorporates a number of mechanisms that afford mobile devices some flexibility in making uplink transmission decisions. With distributed decisions, however, comes the possibility of harmful interactions that could lower overall system efficiency. In this paper, we carry out a preliminary investigation of a proposed token bucket based mechanism for enabling distributed transmission decisions. We explore a class of simple probabilistic policies that readily lend themselves to distributed operation, and show that their efficacy greatly depends on the number of devices sharing the uplink. Specifically, we identify regions in the number of active devices in which both the optimal (in terms of overall system throughput) transmission policy and the device token efficiency (the achieved transmission rate per token) exhibit significantly different properties. Our findings provide initial insight into how to configure token bucket parameters and transmission policies in order to achieve a reasonable compromise between distributed decisions and overall system efficiency.
2009 17th IEEE International Conference on Network Protocols, 2009
Proceedings of the 14th ACM international conference on Mobile computing and networking - MobiCom '08, 2008
We present an experimental characterization of the physical and MAC layers in CDMA 1xEV-DO and th... more We present an experimental characterization of the physical and MAC layers in CDMA 1xEV-DO and their impact on transport layer performance. The 1xEV-DO network is currently the fastest mobile broadband cellular network, offering data rates of up to 3.1 Mbps for both stationary and mobile users. These rates are achieved by using novel capacity enhancement techniques at the lower layers. Specifically, 1xEV-DO incorporates rapid channel rate adaptation in response to signal conditions, and opportunistic scheduling to exploit channel fluctuations. Although shown to perform well in isolation, there is no comprehensive literature that examines the impact of these features on transport layer and application performance in real networks.
This paper investigates the impact of traffic granularity on the performance of routing algorithm... more This paper investigates the impact of traffic granularity on the performance of routing algorithms that incorporate traffic information. Our focus is on robustness of routing performance. The granularity of a traffic matrix can affect network robustness through two opposing factors. Fine granularity traffic matrices allow flexibility in load balancing over a large number of paths by virtue of a large number of streams. This improves the ability of the network to absorb traffic fluctuations. However, as we shall show in this work, increasing the granularity of matrix can increase the variability of the streams which degrades the robustness of a routing solution. Moreover, increasing the granularity of the matrix may also increase the forwarding state cost due to an increase in the number of classifiers and paths. We investigate the trade-off between these two opposing factors to determine the appropriate operating point in terms of traffic granularity. Our analysis is based on traffi...
Proceeding of IEEE International Symposium on a World of Wireless, Mobile and Multimedia Networks 2014, 2014
Lecture Notes in Computer Science, 2007
Modern CDMA wireless channels support multiple transmission rates, which can be dynamically assig... more Modern CDMA wireless channels support multiple transmission rates, which can be dynamically assigned to users based on traffic demand. However, in practice, assignment of high rate channels comes with the penalty of increased power as well as smaller orthogonal codes, which constrains their assignment to only a subset of active users. This motivates the need to carefully control high rate channel assignments so as to minimize power and achieve fairness among users. In this work, we propose a simple class of channel allocation policies to achieve this goal for TCP sessions. We develop an analytical model that explicitly captures both TCP dynamics and the impact of multiple users contending for a shared resource to evaluate the performance of the allocation policy. The model is shown to be accurate by comparing against ns-2 simulations and its utility demonstrated by computing the minimum number of required high rate channels to minimize contention.
Proceedings of the 12th annual international conference on Mobile computing and networking - MobiCom '06, 2006
TCP is the dominant transport protocol over both wired and wireless links. It is however, well kn... more TCP is the dominant transport protocol over both wired and wireless links. It is however, well known that TCP is not suitable for wireless networks and several solutions have been proposed to rectify this shortcoming. In this work we explore cross-layer optimization of the rate adaptation feature of cellular networks to optimize TCP throughput. Modern cellular networks incorporate RF technology that allows them to dynamically vary the wireless channel rate in response to user demand and channel conditions. However, the set of data rates as well as the scheduler policy are typically chosen to optimize throughput for inelastic applications. In order to optimize such a system for TCP, we propose a two state TCP-aware scheduler that switches between two rates as a function of the TCP sending rate. We develop a fluid model of the steady-state TCP behavior for such a system and derive analytical expressions for TCP throughput that explicitly account for rate variability as well as the dependency between the scheduler and TCP. Using the model we choose RF layer parameters that, in conjunction with the TCP-aware scheduler, improve long-term TCP throughput by the order of 15−25%. We also compare our analytical results against those obtained from ns-2 simulations and confirm that our model indeed closely approximates TCP behavior in such an environment.
Proceedings of the third ACM international workshop on Wireless network testbeds, experimental evaluation and characterization - WiNTECH '08, 2008
We describe our experiences in building a city-wide infrastructure for wide-area wireless experim... more We describe our experiences in building a city-wide infrastructure for wide-area wireless experimentation. Our infrastructure has two components -(i) a vehicular testbed consisting of wireless nodes, each equipped with both cellular (EV-DO) and WiFi interfaces, and mounted on city buses plying in Madison, Wisconsin, and (ii) a software platform to utilize these testbed nodes to continuously monitor and characterize performance of large scale wireless networks, such as city-wide mesh networks, unplanned deployments of WiFi hotspots, and cellular networks. Beyond our initial efforts in building and deploying this infrastructure, we have also utilized it to gain some initial understanding of the diversity of user experience in large-scale wireless networks, especially under various mobility scenarios. Since our vehicle-mounted testbed nodes have fairly deterministic mobility patterns, they provide us with much needed performance data on parameters such as RF coverage and available bandwidth, as well as quantify the impact of mobility on performance. We use our initial measurements from this testbed to showcase its ability to provide an efficient, low-cost, and robust method to monitor our target wireless networks. These initial measurements also highlight the challenges we face as we continue to expand this infrastructure. We discuss what these challenges are and how we intend to address them.
2014 IEEE 22nd International Conference on Network Protocols, 2014
Cellular providers are rapidly deploying multiple technologies like cell biasing, carrier aggrega... more Cellular providers are rapidly deploying multiple technologies like cell biasing, carrier aggregation, co-ordinated interference control/scheduling to improve capacity and coverage. In this paper, we explore a complementary transport layer approach based on multipath TCP that can concurrently use multiple interfaces to boost throughput of users with poor coverage and improve fairness. Multipath TCP has been recently standardized by IETF and requires no modifications to applications. It has been shown to improve fairness and throughput in wireline environments and individual user throughputs in wireless networks. However, in a wireless multi-user environment, it is not clear that it is always beneficial, as we show in this paper. Therefore, we examine if it is indeed beneficial for a service provider to judiciously decide whether to enable multiple cellular interfaces on a smartphone based on a global centralized view of its network. Alternatively, should a device decide independently based only on a local view? To quantify the network wide impact in a system where users have multiple cellular interfaces, we have developed centralized and distributed heuristic algorithms to evaluate this, particularly in the context of fairness across all the users. Our simulations and numerical models show that there are potential gains in fairness (15-30%) to be realized by judiciously enabling multipath connections at the cell edge. These gains diminish as the number of users in a cell increases or users behave greedily. We also quantify the delicate balance between throughput and fairness. Our analysis provides an intuition on which user(s) in a cellular network stand to benefit the most by enabling multiple interfaces. We also discuss LTE protocol mechanisms to enforce associations of specific interfaces to specific cells.
2010 IEEE Wireless Communication and Networking Conference, 2010
We consider a number of related problem formulations pertaining to adaptive subcarrier allocation... more We consider a number of related problem formulations pertaining to adaptive subcarrier allocation in multiuser Orthogonal Frequency-Division Multiplexing (OFDM) systems, and prove that they are NP-hard. Thus there exist no known algorithms that can provide optimal solutions for all instances of these problems in polynomial time. We further prove that these problems are hard to approximate in polynomial time. Finally, we discuss qualitatively the settings under which worstcase performance is likely to be observed. If P = NP , then a problem is as easy to compute as to verify it. Consequently, for all the well-known hard NP-complete problems, such as Traveling Salesman, there would would exist an algorithm which can guarantee optimal solutions in polynomial computation time. It is generally conjectured that P = NP . For more details on NP-hardness, see [10].
Proceeding of the 14th ACM SIGKDD international conference on Knowledge discovery and data mining - KDD 08, 2008
Lecture Notes in Computer Science, 2005
An important requirement of a robust traffic engineering solution is insensitivity to changes, be... more An important requirement of a robust traffic engineering solution is insensitivity to changes, be they in the form of traffic fluctuations or changes in the network topology because of link failures. In this paper we focus on developing a fast and effective technique to compute traffic engineering solutions for Interior Gateway Protocol (IGPs) environments that are robust to link failures in the logical topology. The routing and packet forwarding decisions for IGPs is primarily governed by link weights. Our focus is on computing a single set of link weights for a traffic engineering instance that performs well over all single logical link failures. Such types of failures, although usually not long lasting, of the order of tens of minutes, can occur with high enough frequency, of the order of several a day, to significantly affect network performance. The relatively short duration of such failures coupled with issues of computational complexity and convergence time due to the size of current day networks discourage adaptive reactions to such events. Consequently, it is desirable to a priori compute a routing solution that performs well in all such scenarios. Through computational evaluations we demonstrate that our technique yields link weights that perform well over all single link failures and also scales well, in terms of computational complexity, with the size of the network.
IEEE INFOCOM 2014 - IEEE Conference on Computer Communications, 2014
ABSTRACT Proportional Fair (PF) scheduling algorithms are the de-facto standard in cellular netwo... more ABSTRACT Proportional Fair (PF) scheduling algorithms are the de-facto standard in cellular networks. They exploit the users' channel state diversity (induced by fast-fading), and are optimal for stationary channel state distributions and an infinite time-horizon. However, mobile users experience a non-stationary channel, due to slow-fading (on the order of seconds), and are associated with basestations for short periods. Hence, we develop the Predictive Finite-horizon PF Scheduling ((PF)2S) Framework that exploits mobility. We present extensive channel measurement results from a 3G network and characterize mobility-induced channel state trends. We show that a user's channel state is highly reproducible and leverage that to develop a data rate prediction mechanism. We then present a few channel allocation estimation algorithms that rely on the prediction mechanism. Our trace-based simulations consider instances of the PF2S Framework composed of combinations of prediction and channel allocation estimation algorithms. They indicate that the framework can increase the throughput by 15%-55% compared to traditional PF schedulers, while improving fairness.