Recent datacenter network topologies are shifting towards heterogeneous and structured topologies for high throughput, low cost, and simple manageability. However, they rely on sub-optimal routing approaches that fail to achieve their designed capacity. This paper proposes a process for designing optimal oblivious routing that is programmed compactly on programmable switches. The process consists of three contributions in tandem. We first transform a robust optimization problem for designing oblivious routing into a linear program, which is solvable in polynomial time but cannot scale for datacenter topologies. We then prove that the repeated structures in a datacenter topology lead to a structured optimal solution. We use this insight to formulate a scalable linear program, so an optimal oblivious routing solution is obtained in polynomial time for large-scale topologies. For real-world deployment, the optimal solution is converted into forwarding rules for programmable switches with stringent memory. With this constraint, we utilize the repeated structures in the optimal solution to group the forwarding rules, resulting in compact forwarding rules with a much smaller memory requirement. Extensive evaluations show our process i) obtains optimal solutions faster and more scalable than a state-of-the-art technique and ii) reduces the memory requirement by no less than 90% for most considered topologies.

The freshness of real-time status processing of time-sensitive information is crucial for many applications, including flight control, image processing, and autonomous vehicles. In this paper, unprocessed information is sent from sensors to a base station over a shared, unreliable wireless network. The base station has a set of dedicated non-preemptive processors with constant processing times to process information from each sensor. The age of processed information is the time elapsed since the generation of the packet that the processor most recently processed. Our objective is to minimize the expected weighted sum of this age over an infinite time horizon. Here, the challenge is the coupling between a scheduling problem under unreliable communications and the processing times. We first break the coupling by tracking the age of information during processing and derive a lower performance bound of the objective. We then design a stationary randomized policy and a Max-Weight policy for two queueing disciplines: no queues and single-packet queues to achieve our objective. We prove that these policies achieve performance within a factor of two from the optimal. In addition, we prove queues are useful to the stationary randomized policies in highly unreliable or large network settings. Our analytical results are further validated by numerical experiments.

The freshness of real-time status processing of time-sensitive information is crucial for several applications, including patient monitoring and autonomous driving. This freshness is considered in this paper for the system where unprocessed information is sent from sensors to a base station over a shared wireless network. The base station has a dedicated non-preemptive processor with a constant processing time to process information from each sensor. The age of processed information is the time elapsed since the generation of the packet that was most recently processed by a processor. Our objective is to minimize the average age of processed information over an infinite time-horizon. We first show that a drop-free policy simplifies the system without sacrificing optimality. From this simplification, we propose three transmission-scheduling policies with 2-optimal guarantees for different requirements. A distributed Power-2 policy can be implemented without a central scheduler. With a central scheduler, both Back-Off and Max-Weight policies are near optimal with different advantages. The Back-Off policy guarantees a bound on the maximum age of processed information, while the Max-Weight policy achieves the lowest average age in simulation without the guarantee of bound. Simulation results confirm our theoretical findings.

Recent datacenter network topologies are shifting towards heterogeneous and structured topologies for high throughput, low cost, and simple manageability. However, they rely on sub-optimal routing approaches that fail to achieve their designed capacity. This paper proposes a process for designing optimal oblivious routing that is programmed compactly on programmable switches. The process consists of three contributions in tandem. We first transform a robust optimization problem for designing oblivious routing into a linear program, which is solvable in polynomial time but cannot scale for datacenter topologies. We then prove that the repeated structures in a datacenter topology lead to a structured optimal solution. We use this insight to formulate a scalable linear program, so an optimal oblivious routing solution is obtained in polynomial time for large-scale topologies. For real-world deployment, the optimal solution is converted into forwarding rules for programmable switches with stringent memory. With this constraint, we utilize the repeated structures in the optimal solution to group the forwarding rules, resulting in compact forwarding rules with a much smaller memory requirement. Extensive evaluations show our process i) obtains optimal solutions faster and more scalable than a state-of-the-art technique and ii) reduces the memory requirement by no less than 90% for most considered topologies.

Oblivious routing distributes traffic from sources to destinations following predefined routes with rules independent of traffic demands. While finding optimal oblivious routing with a concave objective is intractable for general topologies, we show that it is tractable for structured topologies often used in datacenter networks. To achieve this, we apply graph automorphism and prove the existence of the optimal automorphism-invariant solution. This result reduces the search space to targeting the optimal automorphism-invariant solution. We design an iterative algorithm to obtain such a solution by alternating between convex optimization and a linear program. The convex optimization finds an automorphism-invariant solution based on representative variables and constraints, making the problem tractable. The linear program generates adversarial demands to ensure the final result satisfies all possible demands. Since the construction of the representative variables and constraints are combinatorial problems, we design polynomial-time algorithms for the construction. We evaluate the iterative algorithm in terms of throughput performance, scalability, and generality over three potential applications. The algorithm i) improves the throughput up to 87.5% for partially deployed FatTree and achieves up to$2.55\times $throughput gain for DRing over heuristic algorithms, ii) scales for three considered topologies with a thousand switches, iii) applies to a general structured topology with non-uniform link capacity and server distribution.

Oblivious routing distributes traffic from sources to destinations following predefined routes with rules independent of traffic demands. While finding optimal oblivious routing is intractable for general topologies, we show that it is tractable for structured topologies often used in datacenter networks. To achieve this, we apply graph automorphism and prove the existence of the optimal automorphism-invariant solution. This result reduces the search space to targeting the optimal automorphism-invariant solution. We design an iterative algorithm to obtain such a solution by alternating between two linear programs. The first program finds an automorphism-invariant solution based on representative variables and constraints, making the problem tractable. The second program generates adversarial demands to ensure the final result satisfies all possible demands. Since, the construction of the representative variables and constraints are combinatorial problems, we design polynomial-time algorithms for the construction. We evaluate proposed iterative algorithm in terms of throughput performance, scalability, and generality over three potential applications. The algorithm i) improves the throughput up to 87.5% over a heuristic algorithm for partially deployed FatTree, ii) scales for FatClique with a thousand switches, iii) is applicable to a general structured topology with non-uniform link capacity and server distribution.

The advancements in Internet-connected drones and edge computing raise the possibility of an on-the-fly inspection service that can instantly report the results. Inspection of sites using a drone fleet requires planning to meet situational requirements while minimizing operational costs under uncertainties of inspection requests, the urgency of reports, and the availability of communication channels. In this work, with the restriction on drone flying time, we decompose the planning into two phases: 1) precomputing groups of sites and 2) three-stage stochastic programming. The former phase generates feasible groups of sites, each of which is served by a drone, and precomputes the minimum-cost flying path for each group. The latter phase, given the feasible groups, jointly optimizes drone placement and data delivery under the uncertainties. The two-phase approach allows the planning to scale up to a practical situation. The performance evaluations show that the overall cost can be saved even if uncertainties exist, and the proposed approach significantly outperforms other methods, which do not consider the uncertainties. For a larger problem size, a heuristic algorithm is proposed to trade a loss in the optimality of 1.05–1.09 times the cost with 3.67–483.97 times speed-up in computation time.

While prior work has explored many proposed datacenter designs, only two designs, Clos-based and expander-based, are generally considered practical because they can scale using commodity switching chips. Prior work has used two different metrics, bisection bandwidth and throughput, for evaluating these topologies at scale. Little is known, theoretically or practically, how these metrics relate to each other. Exploiting characteristics of these topologies, we prove an upper bound on their throughput, then show that this upper bound better estimates worst-case throughput than all previously proposed throughput estimators and scales better than most of them. Using this upper bound, we show that for expander-based topologies, unlike Clos, beyond a certain size of the network, no topology can have full throughput, even if it has full bisection bandwidth; in fact, even relatively small expander-based topologies fail to achieve full throughput. We conclude by showing that using throughput to evaluate datacenter performance instead of bisection bandwidth can alter conclusions in prior work about datacenter cost, manageability, and reliability.

Network Functions (NFs) perform on-path processing of network traffic. ISPs are deploying NF Virtualization (NFV) with software NFs run on commodity servers. ISPs aim to ensure that NF chains, directed acyclic graphs of NFs, do not violate Service Level Objectives (SLOs) promised by the ISP to its customers. To meet SLOs, NFV systems sometimes leverage on-path hardware (such as programmable switches and smart NICs) to accelerate NF execution.

The performance and availability of cloud and content providers often depends on the wide area networks (WANs) they use to interconnect their datacenters. WAN routers, which connect to each other using trunks (bundles of links), are sometimes built using an internal Clos topology connecting merchant-silicon switches. As such, these routers are susceptible to internal link and switch failures, resulting in reduced capacity and low availability. Based on the observation that today's WAN routers use relatively simple trunk wiring and routing techniques, we explore the design of novel wiring and more sophisticated routing techniques to increase failure resilience. Specifically, we describe techniques to 1) optimize trunk wiring to increase effective internal router capacity so as to be resilient to internal failures, 2) compute the effective capacity under different failure patterns, and 3) use these to compute compact routing tables under different failure patterns, since switches have limited routing table sizes. Our evaluations show that our approach can mask failures of up to 75% of switches in some cases without exceeding routing table limits, whereas competing techniques can sometimes lose half of a WAN router's capacity with a single failure.

One practical open problem is the development of a distributed algorithm that achieves near-optimal utility using only a finite (and small) buffer size for queues in a stochastic network. This paper studies utility maximization (or cost minimization) in a finite-buffer regime and considers the corresponding delay and reliability (or rate of packet drops) tradeoff. A floating-queue algorithm allows the stochastic network optimization framework to be implemented with finite buffers at the cost of packet drops. Further, the buffer size requirement is significantly smaller than previous works in this area. With a finite buffer size of B packets, the proposed algorithm achieves within O(e-B) of the optimal utility while maintaining average per-hop delay of O(B) and an average per-hop drop rate of O(e-B) in steady state. From an implementation perspective, the floating-queue algorithm requires little modification of the well-known Drift-Plus-Penalty policy (including MaxWeight and Backpressure policies). As a result, the floating-queue algorithm inherits the distributed and low complexity nature of these policies.

This paper considers time-average stochastic optimization, where a time average decision vector, an average of decision vectors chosen in every time step from a time-varying (possibly non-convex) set, minimizes a convex objective function and satisfies convex constraints. This formulation has applications in networking and operations research. In general, time-average stochastic optimization can be solved by a Lyapunov optimization technique. This paper shows that the technique exhibits a transient phase and a steady state phase. When the problem has a unique vector of Lagrange multipliers, the convergence time can be improved. By starting the time average in the steady state, the convergence times become O(1/ε) under a locally-polyhedral assumption and O(1/ε1.5) under a locally-non-polyhedral assumption, where e denotes the proximity to the optimal objective cost.

An information collection problem in a wireless network with random events is considered. Wireless devices report on each event using one of multiple reporting formats. Each format has a different quality and uses different data lengths. Delivering all data in the highest-quality format can overload system resources. The goal is to make intelligent format selection and routing decisions to maximize time-averaged information quality subject to network stability. Lyapunov optimization theory can be used to solve such a problem by repeatedly minimizing the linear terms of a quadratic drift-plus-penalty expression. To reduce delays, this paper proposes a novel extension of this technique that preserves the quadratic nature of the drift minimization while maintaining a fully separable structure. In addition, to avoid high queuing delay, paths are restricted to at most 2 hops. The resulting algorithm can push average information quality arbitrarily close to optimum, with a tradeoff in queue backlog. The algorithm compares favorably to the basic drift-plus-penalty scheme in terms of backlog and delay.

An information collection problem in a wireless network with random events is considered. Wireless nodes report on each event using one of multiple reporting formats. Each format has a different quality and uses a different number of bits. Delivering all data in the highest quality format can overload system resources. The goal is to make intelligent format selection and routing decisions to maximize time-averaged information quality subject to network stability. Lyapunov optimization theory can be used to solve such a problem by repeatedly minimizing the linear terms of a quadratic drift-plus-penalty expression. To reduce delays, a novel extension of this technique that preserves the quadratic nature of the drift minimization while maintaining a separable decision structure is proposed. Also, paths are restricted to 1 or 2 hops to avoid high queuing delay. The resulting algorithm can push average information quality arbitrarily close to optimum, with a trade-off in average delay. The algorithm compares favorably to the basic drift-pluspenalty scheme in terms of backlog and delay.

In wireless multi-hop networks, the allocation of resources is influenced by mechanisms for medium access control (MAC), routing, congestion control, and flow control. Designing these mechanisms jointly can increase the capacity of wireless networks. We attempt to introduce routing into an existing framework for the joint design of flow control and MAC on random access multi-hop wireless networks. The problem of joint flow control, routing and MAC in random access multi-hop wireless network is formulated as an optimization problem. However, a direct formulation yields a non-convex optimization problem. To overcome the difficulty in solving a non-convex problem, we introduce a harmonic rate function to convexify the formulation. The joint optimization mechanism is presented as an iterative process to compute a solution. The resultant distributed algorithm is proved to yield a global optimal solution when it converges. Numerical results are provided to show the convergence of the proposed algorithm and properties of the harmonic rate function.