Network security is widely recognized as a key application of quantum technology. However, its large-scale deployment is hindered by the need for tight coordination between fundamentally different quantum and classical processing steps in conventional protocols. This requirement introduces strong cross-layer interdependencies that conflict with the modular, layered architectures enabling scalability in modern communication networks. Here, we present an alternative strategy that confines all quantum interventions to the physical layer and remains transparently compatible with existing network abstractions. This is achieved by directly embedding quantum features and classical information within the same optical field using bright squeezed light. Physical-layer signals are analyzed using a cumulative sum (CUSUM) method to enable quantum-enhanced threat detection. We validate the practicality of this approach through field deployment over a metropolitan-scale fiber network and further demonstrate network-level security functionalities enabled by physical-layer quantum-enhanced thread detection. These results establish a practical, scalable framework for seamlessly integrating quantum-enhanced security into large-scale communication infrastructure.

Augmenting a train of bright phase-modulated laser-light pulses of a coherent communications system with infinitesimally small quantum photons per pulse-entangled across several time bins-prepared by splitting squeezed light in a temporal-mode interferometer can dramatically enhance a homodyne receiver's ability to detect a sudden change in the channel loss, by up to a factor that is the inverse of the pre-change loss, without affecting the communications rate. We discuss the quantum limit of quickest change detection, and the problem of joint communications and change detection that our Letter opens up.

We consider the problem of distilling efficient network topologies for collective communications. We provide an algorithmic framework for constructing direct-connect topologies optimized for the latency vs. bandwidth trade-off associated with the workload. Our approach synthesizes many different topologies and communication schedules for a given cluster size and degree, then identifies the best option for a given workload. Our algorithms start from small, optimal base topologies and associated schedules, using techniques that can be iteratively applied to derive much larger topologies and schedules. Additionally, we incorporate well-studied large-scale graph topologies into our algorithmic framework by producing efficient communication schedules for them using a novel polynomial-time algorithm. Our evaluation uses multiple testbeds and large-scale simulations to demonstrate significant performance benefits from our derived topologies and schedules.

The ability to localize transmission loss change to a subset of links in optical networks is crucial for maintaining network reliability, performance and security. Quantum probes, implemented by sending blocks of n coherent-state pulses augmented with continuous-variable (CV) squeezing (n=1) or weak temporal-mode entanglement (n>1) over a lossy channel to a receiver with homodyne detection capabilities, are known to be more sensitive than their quasi-classical counterparts in detecting a sudden increase in channel loss. The enhanced sensitivity can be characterized by the increased Kullback-Leibler (KL) divergence of the homodyne output, before and after the loss change occurs. When combined with the theory of quickest change detection (QCD), the increase in KL divergence translates into a decrease in detection latency. In this work, we first revisit quantum probes over a channel, generalizing previous results on n=1 (CV squeezed states) to arbitrary values of n. Assuming a subset of nodes in an optical network is capable of sending and receiving such probes through intermediate nodes with all-optical switching capabilities, we present a scheme for quickly detecting the links that have suffered a sudden drop in transmissivity. Since quantum probes lose their sensitivity with increasing loss in the channel, we first propose a probe construction algorithm that makes the set of links suffering transmission loss change identifiable, while minimizing the longest distance a probe traverses. We then introduce new cumulative sum (CUSUM) statistics with a stopping rule, which allows us to run the CUSUM algorithm to quickly localize the lossy links using our constructed probes. Finally, we show that the proposed scheme achieves a quantum speedup in decreasing the detection delay.

The all-to-all collective communications primitive is widely used in machine learning (ML) and high performance computing (HPC) workloads, and optimizing its performance is of interest to both ML and HPC communities. All-to-all is a particularly challenging workload that can severely strain the underlying interconnect bandwidth at scale. This paper takes a holistic approach to optimize the performance of all-to-all collective communications on supercomputer-scale direct-connect interconnects. We address several algorithmic and practical challenges in developing efficient and bandwidth-optimal all-to-all schedules for any topology and lowering the schedules to various runtimes and interconnect technologies. We also propose a novel topology that delivers near-optimal all-to-all performance.

The growing demand for edge computing resources, particularly due to increasing popularity of Internet of Things (IoT), and distributed machine/deep learning applications poses a significant challenge. On the one hand, certain edge service providers (ESPs) may not have sufficient resources to satisfy their applications according to the associated service-level agreements. On the other hand, some ESPs may have additional unused resources. In this paper, we propose a resource-sharing framework that allows different ESPs to optimally utilize their resources and improve the satisfaction level of applications subject to constraints such as communication cost for sharing resources across ESPs. Our framework considers that different ESPs have their own objectives for utilizing their resources, thus resulting in a multi-objective optimization problem. We present an N-person Nash Bargaining Solution (NBS) for resource allocation and sharing among ESPs with Pareto optimality guarantee. Furthermore, we propose a distributed, primal-dual algorithm to obtain the NBS by proving that the strong-duality property holds for the resultant resource sharing optimization problem. Using synthetic and real-world data traces, we show numerically that the proposed NBS based framework not only enhances the ability to satisfy applications' resource demands, but also improves utilities of different ESPs.

Throughout history, central squares and main pedestrian arteries of cities provided goods and services to meet the needs of their catchment areas. This study examines which types of businesses tend to agglomerate together and which ones are spread out. In addition, we analyzed whether there is a correlation between the location of the businesses and their clientele. Our research question is whether a Machine Learning (ML) system be trained to predict the best-performing locations for new commercial uses in the city—using city data showing openings/closures of various types of business during the last ten years. This study is based on four main phases: (1) Geoeconomic analysis by calculating ordered good index, (2) Clustering business categories by geolocation features, (3) Clustering business categories by the Wasserstein metric, and (4) Developing an interface for a ML predictor. The performance of the ML setup improves significantly when including additional features like demographic and socio-economic criteria in a supervised deep learning environment. This study suggests that our approach can help cities identify apt locations for a particular business type. We believe our approach can be applied to the disposition of public and government building programs in the future, which can dramatically help cities optimize their urban program layouts.

Distributed load balancing is the act of allocating jobs among a set of servers as evenly as possible. The static interpretation of distributed load balancing leads to formulating the load-balancing problem as a classical balls-and-bins problem with jobs (balls) never leaving the system and accumulating at the servers (bins). While most of the previous work in the static setting focus on studying the maximum number of jobs allocated to a server or maximum load , little importance has been given to the implementation cost , or the cost of moving a job/data to/from its allocated server, for such policies. This article designs and evaluates server proximity aware static load-balancing policies with a goal to reduce the implementation cost . We consider a class of proximity aware Power of Two (POT) choice-based assignment policies for allocating jobs to servers, where both jobs and servers are located on a two-dimensional Euclidean plane. In this framework, we investigate the tradeoff between the implementation cost and load-balancing performance of different allocation policies. To this end, we first design and evaluate a Spatial Power of two (sPOT) policy in which each job is allocated to the least loaded server among its two geographically nearest servers. We provide expressions for the lower bound on the asymptotic expected maximum load on the servers and prove that sPOT does not achieve classical POT load-balancing benefits. However, experimental results suggest the efficacy of sPOT with respect to expected implementation cost. We also propose two non-uniform server sampling-based POT policies that achieve the best of both implementation cost and load-balancing performance. We then extend our analysis to the case where servers are interconnected as an n -vertex graph G(S, E) . We assume each job arrives at one of the servers, u, chosen uniformly at random from the vertex set S. We then assign each job to the server with minimum load among servers u and v where v is chosen according to one of the following two policies: (i) Unif-POT( k ): Sample a server v uniformly at random from k -hop neighborhood of u; (ii) InvSq-POT( k ): Sample a server v from k -hop neighborhood of u with probability proportional to the inverse square of the distance between u and v . An extensive simulation over a wide range of topologies validates the efficacy of both the policies. Our simulation results show that both policies consistently produce a load distribution that is much similar to that of a classical POT. Depending on topology, we observe the total variation distance to be of the order of 0.002–0.08 for both the policies while achieving a 8%–99% decrease in implementation cost as compared to the classical POT.

We consider the problem of distilling optimal network topologies for collective communications. We provide an algorithmic framework for constructing direct-connect topologies optimized for the latency-bandwidth tradeoff given a collective communication workload. Our algorithmic framework allows us to start from small base topologies and associated communication schedules and use a set of techniques that can be iteratively applied to derive much larger topologies and associated schedules. Our approach allows us to synthesize many different topologies and schedules for a given cluster size and degree constraint, and then identify the optimal topology for a given workload. We provide an analytical-model-based evaluation of the derived topologies and results on a small-scale optical testbed that uses patch panels for configuring a topology for the duration of an application’s execution. We show that the derived topologies and schedules provide significant performance benefits over existing collective communications implementations.

The proliferation of smart devices, computational and storage resources is predicted to continue aggressively in the near future. Such “networked” devices and resources which are distributed in a physical space and provide services are collectively referred to as a distributed service network . Assigning users or applications to available resources is important to sustain high performance of the distributed service network. In this work, we consider a one-dimensional service network where both users and resources are located on a line, and analyze a unidirectional assignment policy Move To Right (MTR), which sequentially assigns users to resources available to their right. We express the communication cost for a user-resource assignment as an increasing function of the distance traveled by the user request ( request distance ) and analyze the expected communication cost for the service network when locations of users and resources are modeled by different spatial point processes. We use results from the literature that map the request distance of an assigned user in a one-dimensional service network to the sojourn time of a customer in an exceptional service accessible batch queueing system. We compute the Laplace–Stieltjes transform of the sojourn time distribution for this queueing system for Poisson distributed users with general inter-resource distance distributions and in the process also generate new results for batch service queues. Unlike previous work (Panigrahy et al. in Perform Eval 142:102, 2020), our framework not only captures the first-order moment of the request distance, but also the request distance distribution itself, thus allowing us to compute the expected communication cost under different cost models.

Machine learning (ML) has popular applications in domains involving image, video, text and voice. However, in architecture, image-based ML systems face challenges capturing the complexity of three-dimensional space. In this paper, we leverage a graph-based ML system that can capture the inherent topology of architectural conceptual designs and identify high-performing latent patterns within such designs. In particular, our goal is to translate architectural graph data into three-dimensional massing models. We are building on our prior ML work, where we, a. discovered latent topological features, b. composed building blocks into new designs, c. evaluated their feasibility, and d. explored Generative Adversarial (Neural) Networks (GAN)-generated design variations. We trained the ML system with architectural design data that we gathered from an online architectural design competition platform, translated them into machine-readable graph representations, and identified their essential subgraphs to develop novel compositions. In this paper, we explore how these novel designs (outputted in graph form), can be translated into three-dimensional architectural form. We present an ML approach to turn graph representations into functional volumetric massing models. The ultimate goal of the study is to develop an end-to-end pipeline to generate architectural design - from a graph representation to a fully developed conceptual proxy of a designed product. The research question is promising in automating conceptual design, and we believe the outcome can be relevant to other design disciplines as well.

Exposing architectural concepts to modern AI/ML methods requires succinct representations that can capture the essential relationship between adjacent architectural elements in design samples. We explore various examples of topological spaces—from rudimentary graphs to richer closure-finite weak topology (CW)-complexes—and how they can be used to model architectural design samples. We then argue how AI concepts developed for graphs can be naturally extended to more general topological spaces with the following goals in mind: (1) finding salient building-block substructures that occur consistently in highly rated designs; (2) unearthing latent design rules from data by characterizing the space of designs using graph representation learning followed by topological dimensionality reduction; and (3) generating novel designs—given training design samples—using generative adversarial networks applied to graphs. Our preliminary results on the application of AI/ML on graphs are based on the architectural design data from Arcbazar, an online crowdsourcing platform for architectural designs.

Providing resources to different users or applications is fundamental to cloud computing. This is a challenging problem as a cloud service provider may have insufficient resources to satisfy all user requests. Furthermore, allocating available resources optimally to different applications is also challenging. Resource sharing among different cloud service providers can improve resource availability and resource utilization as certain cloud service providers may have free resources available that can be "rented" by other service providers. However, different cloud service providers can have different objectives or utilities. Therefore, there is a need for a framework that can share and allocate resources in an efficient and effective way, while taking into account the objectives of various service providers that results in a multi-objective optimization problem. In this paper, we present a Cooperative Game Theory (CGT) based framework for resource sharing and allocation among different service providers with varying objectives that form a coalition. We show that the resource sharing problem can be modeled as an N-player canonical cooperative game with non-transferable utility (NTU) and prove that the game is convex for monotonic non-decreasing utilities. We propose an O(N) algorithm that provides an allocation from the core, hence guaranteeing Pareto optimality. We evaluate the performance of our proposed resource sharing framework in a number of simulation settings and show that our proposed framework improves user satisfaction and utility of service providers.

This chapter discusses various use cases of artificial intelligence (AI) in crowdsourcing architecture. Online platforms, such as http://Arcbazar.com">Arcbazar.com, are based on the traditional competition model and harness design intelligence from an ever-growing global designer community. We give a brief background on competitions, take a closer look on crowdsourcing, discuss competition vs collaboration models, demonstrate how AI has been used in various aspects of the crowdsourcing protocol, such as price recommendation and space-surveying systems, and deliberate on the potential integration of AI in generating conceptual designs for future crowdsourced projects. The use of AI in architectural design acquisition can offer immense opportunities to expand the reach of competitive design solutions to a wider global audience. We end with a speculative discussion of how this nascent transformation in architecture can affect the role of the architect, and the future of architectural practice.

Distributed load balancing is the act of allocating jobs among a set of servers as evenly as possible. There are mainly two versions of the load balancing problem that have been studied in the literature: static and dynamic. The static interpretation leads to formulating the load balancing problem as a case with jobs (balls) never leaving the system and accumulating at the servers (bins) whereas the dynamic setting deals with the case when jobs arrive and leave the system after service completion. This paper designs and evaluates server proximity aware job allocation policies for treating load balancing problems with a goal to reduce the communication cost associated with the jobs. We consider a class of proximity aware Power of Two (POT) choice based assignment policies for allocating jobs to servers, where servers are interconnected as an n-vertex graph G(V, E). For the static version, we assume each job arrives at one of the servers, u. For the dynamic setting, we assume G to be a circular graph and job arrival process at each server is described by a Poisson point process with the job service time exponentially distributed. For both settings, we then assign each job to the server with minimum load among servers u and v where v is chosen according to one of the following two policies: (i) Unif-POT(k): Sample a server v uniformly at random from k-hop neighborhood of u (ii) InvSq-POT(k): Sample a server v from k-hop neighborhood of u with probability proportional to the inverse square of the distance between u and v. Our simulation results show that both the policies consistently produce a load distribution which is much similar to that of a classical proximity oblivious POT policy.

Finding optimal matchings in dense graphs is of general interest and of particular importance in social, transportation and biological networks. While developing optimal solutions for various matching problems is important, the running times of the fastest available optimal matching algorithms are too costly. However, when the vertices of the graphs are point-sets in R^d and edge weights correspond to the euclidean distances, the available optimal matching algorithms are substantially faster. In this paper, we propose a novel network embedding based heuristic algorithm to solve various matching problems in dense graphs. In particular, using existing network embedding techniques, we first find a low dimensional representation of the graph vertices in R^d and then run faster available matching algorithms on the embedded vertices. To the best of our knowledge, this is the first work that applies network embedding to solve various matching problems. Experimental results validate the efficacy of our proposed algorithm.

We address energy-efficient placement of data and analytics components of composite analytics services on a wireless network to minimize execution-time energy consumption (computation and communication) subject to compute, storage and network resource constraints. We introduce an expressive analytics service hypergraph model for representing k-ary composability relationships (k ≥ 2) between various analytics and data components and leverage binary quadratic programming (BQP) to minimize the total energy consumption of a given placement of the analytics hypergraph nodes on the network subject to resource availability constraints. Then, after defining a potential energy functional Φ(·) to model the affinities of analytics components and network resources using analogs of attractive and repulsive forces in physics, we propose a decentralized Metropolis Monte Carlo (MMC) sampling method which seeks to minimize Φ by moving analytics and data on the network. Although Φ is non-convex, using a potential game formulation, we identify conditions under which the algorithm provably converges to a local minimum energy equilibrium placement configuration. Trace-based simulations of the placement of a deep-neural-network analytics service on a realistic wireless network show that for smaller problem instances our MMC algorithm yields placements with total energy within a small factor of BQP and more balanced workload distributions; for larger problems, it yields low-energy configurations while the BQP approach fails.

We consider a class of power of two choice based assignment policies for allocating users to servers, where both users and servers are located on a two-dimensional Euclidean plane. In this framework, we investigate the inherent tradeoff between the communication cost, and load balancing performance of different allocation policies. To this end, we first design and evaluate a Spatial Power of two (sPOT) policy in which each user is allocated to the least loaded server among its two geographically nearest servers sequentially. When servers are placed on a two-dimensional square grid, sPOT maps to the classical Power of two (POT) policy on the Delaunay graph associated with the Voronoi tessellation of the set of servers. We show that the associated Delaunay graph is 4-regular and provide expressions for asymptotic maximum load using results from the literature. For uniform placement of servers, we map sPOT to a classical balls and bins allocation policy with bins corresponding to the Voronoi regions associated with the second order Voronoi diagram of the set of servers. We provide expressions for the lower bound on the asymptotic expected maximum load on the servers and prove that sPOT does not achieve POT load balancing benefits. However, experimental results suggest the efficacy of sPOT with respect to expected communication cost. Finally, we propose two non-uniform server sampling based POT policies that achieve the best of both the performance metrics. Experimental results validate the effectiveness of our proposed policies.

Remote quantum entanglement can enable numerous applications including distributed quantum computation, secure communication, and precision sensing. We consider how a quantum network—nodes equipped with limited quantum processing capabilities connected via lossy optical links—can distribute high-rate entanglement simultaneously between multiple pairs of users. We develop protocols for such quantum “repeater” nodes, which enable a pair of users to achieve large gains in entanglement rates over using a linear chain of quantum repeaters, by exploiting the diversity of multiple paths in the network. Additionally, we develop repeater protocols that enable multiple user pairs to generate entanglement simultaneously at rates that can far exceed what is possible with repeaters time sharing among assisting individual entanglement flows. Our results suggest that the early-stage development of quantum memories with short coherence times and implementations of probabilistic Bell-state measurements can have a much more profound impact on quantum networks than may be apparent from analyzing linear repeater chains. This framework should spur the development of a general quantum network theory, bringing together quantum memory physics, quantum information theory, quantum error correction, and computer network theory.