arXiv subjects

Amr Rizk

Publications and source records attributed to Amr Rizk.

At least 19 recordsLinked to original sources

arXiv2026

Balancing Quantum Memories in Asymmetric Repeaters for High-Fidelity Entanglement Distribution

At the core of the quantum Internet lie quantum repeaters that enable remote end-to-end entanglement generation. Fundamentally, the entanglement generation rate and fidelity of quantum repeaters constitute the bottleneck for end-to-end performance. To achieve high rates, quantum repeaters employ quantum memory multiplexing. In a high-rate standard repeater, each memory sequentially generates an entanglement with its neighboring nodes and then applies entanglement swapping. This, however, results in low fidelity due to decoherence of the first-formed entanglement in the sequential generation process. By allocating different numbers of memories to simultaneously form entanglements with the left and right adjacent nodes, quantum repeaters reduce high waiting times and achieve high fidelity. In such a repeater, a mismatch problem arises due to the difference between the probabilistic number of generated entanglements on both sides. Consequently, some entanglements remain stored until opposite entanglements are available. The mismatch problem reduces the repeater rate and particularly the entanglement fidelity. In this paper, we consider the mismatch problem in an asymmetric repeater with different distances to its adjacent nodes. To mitigate the mismatch problem, we derive a dynamic optimal memory allocation. Under the optimal allocation, we derive statistical lower bounds on the achievable rate and fidelity. We demonstrate that the optimal allocation significantly improves the fidelity while maintaining a comparable rate to the standard repeater. In contrast, our results show that fixed memory allocation may be detrimental to the fidelity.

quant-ph↗

arXiv2026

Analyzing Symbolic Properties for DRL Agents in Systems and Networking

Deep reinforcement learning (DRL) has shown remarkable performance on complex control problems in systems and networking, including adaptive video streaming, wireless resource management, and congestion control. For safe deployment, however, it is critical to reason about how agents behave across the range of system states they encounter in practice. Existing verification-based methods in this domain primarily focus on point properties, defined around fixed input states, which offer limited coverage and require substantial manual effort to identify relevant input-output pairs for analysis. In this paper, we study symbolic properties, that specify expected behavior over ranges of input states, for DRL agents in systems and networking. We present a generic formulation for symbolic properties, with monotonicity and robustness as concrete examples, and show how they can be analyzed using existing DNN verification engines. Our approach encodes symbolic properties as comparisons between related executions of the same policy and decomposes them into practically tractable sub-properties. These techniques serve as practical enablers for applying existing verification tools to symbolic analysis. Using our framework, diffRL, we conduct an extensive empirical study across three DRL-based control systems, adaptive video streaming, wireless resource management, and congestion control. Through these case studies, we analyze symbolic properties over broad input ranges, examine how property satisfaction evolves during training, study the impact of model size on verifiability, and compare multiple verification backends. Our results show that symbolic properties provide substantially broader coverage than point properties and can uncover non-obvious, operationally meaningful counterexamples, while also revealing practical solver trade-offs and limitations.

cs.NI↗

arXiv2026

Scalable On-the-fly Transcoding for Adaptive Streaming of Dynamic Point Clouds

On-the-fly transcoding of dynamic point cloud sequences reduces storage requirements and virtually increases the number of available representations for on demand streaming scenarios. On-the-fly transcoding introduces, however, additional workload to media providers' infrastructure. While V-PCC encoded content can be efficiently transcoded by re-encoding the underlying video bitstreams, which greatly benefits from hardware-accelerated video codec implementations, the scalability of such a system remains unclear. In this work, we introduce and evaluate a dynamic point cloud streaming system that utilizes on-the-fly transcoding. We explore the limits of scalability of this system in terms of request fulfillment times, specifically evaluating the perceived user Quality of Experience. We empirically show how caching and speculative transcoding allow to significantly reduce transcoding loads, allowing to scale to a higher number of simultaneous clients.

cs.MM↗

arXiv2026

Point Cloud Streaming with Latency-Driven Implicit Adaptation using MoQ

Point clouds are a promising video representation for virtual and augmented reality. Their high-bitrate, however, has so far limited the practicality of live streaming systems. In this work, we leverage the delivery timeout feature within the Media Over QUIC protocol to perform implicit server-side adaptation based on an application's latency target. Through experimentation with several publisher and network configurations, we demonstrate that our system unlocks a unique trade-off on a per-client basis: applications with lower latency requirements will receive lower-quality video, while applications with more relaxed latency requirements will receive higher-quality video.

cs.MM↗

arXiv2025

Shesha: Opportunistic In-network Acceleration of Asynchronous Distributed Reinforcement Learning

Large-scale training for distributed Machine Learning can cause congestion at bottleneck switch ports, leading to model staleness through update losses. This is particularly detrimental for asynchronous Distributed Reinforcement Learning (DRL) training, as stale updates are known to degrade convergence performance in asynchronous settings. This paper presents \textit{Shesha}, an in-network DRL accelerator engine, which \textit{opportunistically aggregates} asynchronously generated model updates \textit{on the fly} while they traverse the data plane queue. This aggregation operation motivates an alternative queue design, which we prototype and envision for future Top-of-Rack switches. We further present corresponding host-side transmission control in the face of possible congestion, taking advantage of in-network accelerator feedback. A quantification of model staleness, denoted Age-of-Model (AoM), together with a formal verifier allows us to reason on system-wide AoM objectives in multi DRL-cluster scenarios. Shesha shows significant reductions in model staleness and queue congestion, improving overall convergence behavior for asynchronous DRL workloads.

cs.NI↗

arXiv2024

Stochastic Analysis of Entanglement-assisted Quantum Communication Channels

We present a queueing model for a quantum communication network consisting of a primary queue and a service queue in which Bell pairs are formed and stored. The Bell pairs are inherently extremely short-lived rendering the service queue (the quantum queue) much faster than the primary queue. We study the asymptotic behaviour of this multi-scale queueing system via a stochastic averaging principle. We prove a Functional Law of Large Numbers (FLLN) and a Functional Central Limit Theorem (FCLT) for the standard queue averaging the dynamics of the fast service queue.

math.PR↗

arXiv2024

Learned Compression of Point Cloud Geometry and Attributes in a Single Model through Multimodal Rate-Control

Point cloud compression is essential to experience volumetric multimedia as it drastically reduces the required streaming data rates. Point attributes, specifically colors, extend the challenge of lossy compression beyond geometric representation to achieving joint reconstruction of texture and geometry. State-of-the-art methods separate geometry and attributes to compress them individually. This comes at a computational cost, requiring an encoder and a decoder for each modality. Additionally, as attribute compression methods require the same geometry for encoding and decoding, the encoder emulates the decoder-side geometry reconstruction as an input step to project and compress the attributes. In this work, we propose to learn joint compression of geometry and attributes using a single, adaptive autoencoder model, embedding both modalities into a unified latent space which is then entropy encoded. Key to the technique is to replace the search for trade-offs between rate, attribute quality and geometry quality, through conditioning the model on the desired qualities of both modalities, bypassing the need for training model ensembles. To differentiate important point cloud regions during encoding or to allow view-dependent compression for user-centered streaming, conditioning is pointwise, which allows for local quality and rate variation. Our evaluation shows comparable performance to state-of-the-art compression methods for geometry and attributes, while reducing complexity compared to related compression methods.

cs.CV↗

arXiv2024

Utility-driven Optimization of TTL Cache Hierarchies under Network Delays

We optimize hierarchies of Time-to-Live (TTL) caches under random network delays. A TTL cache assigns individual eviction timers to cached objects that are usually refreshed upon a hit where upon a miss the object requires a random time to be fetched from a parent cache. Due to their object decoupling property, TTL caches are of particular interest since the optimization of a per-object utility enables service differentiation. However, state-of-the-art exact TTL cache optimization does not extend beyond single TTL caches, especially under network delays. In this paper, we leverage the object decoupling effect to formulate the non-linear utility maximization problem for TTL cache hierarchies in terms of the exact object hit probability under random network delays. We iteratively solve the utility maximization problem to find the optimal per-object TTLs. Further, we show that the exact model suffers from tractability issues for large hierarchies and propose a machine learning approach to estimate the optimal TTL values for large systems. Finally, we provide numerical and data center trace-based evaluations for both methods showing the significant offloading improvement due to TTL optimization considering the network delays.

cs.NI↗

arXiv2023

Collaborative Optimization of the Age of Information under Partial Observability

The significance of the freshness of sensor and control data at the receiver side, often referred to as Age of Information (AoI), is fundamentally constrained by contention for limited network resources. Evidently, network congestion is detrimental for AoI, where this congestion is partly self-induced by the sensor transmission process in addition to the contention from other transmitting sensors. In this work, we devise a decentralized AoI-minimizing transmission policy for a number of sensor agents sharing capacity-limited, non-FIFO duplex channels that introduce random delays in communication with a common receiver. By implementing the same policy, however with no explicit inter-agent communication, the agents minimize the expected AoI in this partially observable system. We cater to the partial observability due to random channel delays by designing a bootstrap particle filter that independently maintains a belief over the AoI of each agent. We also leverage mean-field control approximations and reinforcement learning to derive scalable and optimal solutions for minimizing the expected AoI collaboratively.

cs.MA↗

arXiv2023

On the Fidelity Distribution of Link-level Entanglements under Purification

Quantum entanglement is the key to quantum communications over considerable distances. The first step for entanglement distribution among quantum communication nodes is to generate link-level Einstein-Podolsky-Rosen (EPR) pairs between adjacent communication nodes. EPR pairs may be continuously generated and stored in a few quantum memories to be ready for utilization by quantum applications. A major challenge is that qubits suffer from unavoidable noise due to their interaction with the environment, which is called decoherence. This decoherence results in the known exponential decay model of the fidelity of the qubits with time, thus, limiting the lifetime of a qubit in a quantum memory and the performance of quantum applications. In this paper, we evaluate the fidelity of the stored EPR pairs under two opposite dynamical and probabilistic phenomena, first, the aforementioned decoherence and second purification, i.e. an operation to improve the fidelity of an EPR pair at the expense of sacrificing another EPR pair. Instead of applying the purification as soon as two EPR pairs are generated, we introduce a Purification scheme Beyond the Generation time (PBG) of two EPR pairs. We analytically show the probability distribution of the fidelity of stored link-level EPR pairs in a system with two quantum memories at each node allowing a maximum of two stored EPR pairs. In addition, we apply a PBG scheme that purifies the two stored EPR pairs upon the generation of an additional one. We finally provide numerical evaluations of the analytical approach and show the fidelity-rate trade-off of the considered purification scheme.

quant-ph↗

arXiv2023

Accelerating End-host Congestion Response using P4 Programmable Switches

Transport layer congestion control relies on feedback signals that travel from the congested link to the receiver and back to the sender. This forward congestion control loop, first, requires at least one rount-trip time (RTT) to react to congestion and secondly, it depends on the downstream path after the bottleneck. The former property leads to a reaction time in the order of RTT + bottleneck queue delay, while the second may amplify the unfairness due to heterogeneous RTT. In this paper, we present Reverse Path Congestion Marking (RPM) to accelerate the reaction to network congestion events without changing the end-host stack. RPM decouples the congestion signal from the downstream path after the bottleneck while maintaining the stability of the congestion control loop. We show that RPM improves throughput fairness for RTT-heterogeneous TCP flows as well as the flow completion time, especially for small Data Center TCP (DCTCP) flows. Finally, we show RPM evaluation results in a testbed built around P4 programmable ASIC switches.

cs.NI↗

arXiv2023

Optimal Decision Making in Active Queue Management

Active Queue Management (AQM) aims to prevent bufferbloat and serial drops in router and switch FIFO packet buffers that usually employ drop-tail queueing. AQM describes methods to send proactive feedback to TCP flow sources to regulate their rate using selective packet drops or markings. Traditionally, AQM policies relied on heuristics to approximately provide Quality of Service (QoS) such as a target delay for a given flow. These heuristics are usually based on simple network and TCP control models together with the monitored buffer filling. A primary drawback of these heuristics is that their way of accounting flow characteristics into the feedback mechanism and the corresponding effect on the state of congestion are not well understood. In this work, we show that taking a probabilistic model for the flow rates and the dequeueing pattern, a Semi-Markov Decision Process (SMDP) can be formulated to obtain an optimal packet-dropping policy. This policy-based AQM, named PAQMAN, takes into account a steady-state model of TCP and a target delay for the flows. Additionally, we present an inference algorithm that builds on TCP congestion control in order to calibrate the model parameters governing underlying network conditions. Using simulation, we show that the prescribed AQM yields comparable throughput to state-of-the-art AQM algorithms while reducing delays significantly.

cs.PF↗

arXiv2022

Load Balancing in Compute Clusters with Delayed Feedback

Load balancing arises as a fundamental problem, underlying the dimensioning and operation of many computing and communication systems, such as job routing in data center clusters, multipath communication, Big Data and queueing systems. In essence, the decision-making agent maps each arriving job to one of the possibly heterogeneous servers while aiming at an optimization goal such as load balancing, low average delay or low loss rate. One main difficulty in finding optimal load balancing policies here is that the agent only partially observes the impact of its decisions, e.g., through the delayed acknowledgements of the served jobs. In this paper, we provide a partially observable (PO) model that captures the load balancing decisions in parallel buffered systems under limited information of delayed acknowledgements. We present a simulation model for this PO system to find a load balancing policy in real-time using a scalable Monte Carlo tree search algorithm. We numerically show that the resulting policy outperforms other limited information load balancing strategies such as variants of Join-the-Most-Observations and has comparable performance to full information strategies like: Join-the-Shortest-Queue, Join-the-Shortest-Queue(d) and Shortest-Expected-Delay. Finally, we show that our approach can optimise the real-time parallel processing by using network data provided by Kaggle.

cs.DC↗

arXiv2022

A Palm Calculus Approach to the Distribution of the Age of Information

A key metric to express the timeliness of status updates in latency-sensitive networked systems is the age of information (AoI), i.e., the time elapsed since the generation of the last received informative status message. This metric allows studying a number of applications including updates of sensory and control information in cyber-physical systems and vehicular networks as well as, job and resource allocation in cloud clusters. State-of-the-art approaches to analyzing the AoI rely on queueing models that are composed of one or many queuing systems endowed with service order, e.g., FIFO, LIFO, or last-generated-first-out order. A major difficulty arising in these analysis methods is capturing the AoI under message reordering when the delivery is non-preemptive and non-FIFO, i.e., when messages can overtake each other and the reception of informative messages may obsolete some messages that are underway. In this paper, we derive an exact formulation for the distribution of AoI in non-preemptive, non-FIFO systems where the main ingredients of our analysis are Palm calculus and time inversion. Owing to the rationality of the Laplace-Stieltjes transforms that are used in our approach, we obtain computable exact expressions for the distribution of AoI.

cs.PF↗

arXiv2022

On the Impact of Network Delays on Time-to-Live Caching

We consider Time-to-Live (TTL) caches that tag every object in cache with a specific (and possibly renewable) expiration time. State-of-the-art models for TTL caches assume zero object fetch delay, i.e., the time required to fetch a requested object that is not in cache from a different cache or the origin server. Particularly, in cache hierarchies, this delay has a significant impact on performance metrics such as the object hit probability. Recent work suggests that the impact of the object fetch delay on the cache performance will continue to increase due to the scaling mismatch between shrinking inter-request times (due to higher data center link rates) in contrast to processing and memory access times. In this paper, we analyze tree-based cache hierarchies with random object fetch delays and provide an exact analysis of the corresponding object hit probability. Our analysis allows understanding the impact of random delays and TTLs on cache metrics for a wide class of request stream models characterized through Markov arrival processes. This is expressed through a metric that we denote delay impairment of the hit probability. In addition, we analyze and extend state-of-the-art approximations of the hit probability to take the delay into account. We provide numerical and trace-based simulation-based evaluation results showing that larger TTLs do not efficiently compensate for the detrimental effect of object fetch delays. Our evaluations also show that unlike our exact model the state-of-the-art approximations do not capture the impact of the object fetch delay well especially for cache hierarchies. Surprisingly, we show that the impact of this delay on the hit probability is not monotonic but depends on the request stream properties, as well as, the TTL.

cs.NI↗

arXiv2021

DAG-type Distributed Ledgers via Young-age Preferential Attachment

Distributed Ledger Technologies provide a mechanism to achieve ordering among transactions that are scattered on multiple participants with no prerequisite trust relations. This mechanism is essentially based on the idea of new transactions referencing older ones in a chain structure. Recently, DAG-type Distributed Ledgers that are based on directed acyclic graphs (DAGs) were proposed to increase the system scalability through sacrificing the total order of transactions. In this paper, we develop a mathematical model to study the process that governs the addition of new transactions to the DAG-type Distributed Ledger. We propose a simple model for DAG-type Distributed Ledgers that are obtained from a recursive Young-age Preferential Attachment scheme, i.e. new connections are made preferably to transactions that have not been in the system for very long. We determine the asymptotic degree structure of the resulting graph and show that a forward component of linear size arises if the edge density is chosen sufficiently large in relation to the `young-age preference' that tunes how quickly old transactions become unattractive.

math.PR↗

arXiv2021

P4-CoDel: Experiences on Programmable Data Plane Hardware

Fixed buffer sizing in computer networks, especially the Internet, is a compromise between latency and bandwidth. A decision in favor of high bandwidth, implying larger buffers, subordinates the latency as a consequence of constantly filled buffers. This phenomenon is called Bufferbloat. Active Queue Management (AQM) algorithms such as CoDel or PIE, designed for the use on software based hosts, offer a flow agnostic remedy to Bufferbloat by controlling the queue filling and hence the latency through subtle packet drops. In previous work, we have shown that the data plane programming language P4 is powerful enough to implement the CoDel algorithm. While legacy software algorithms can be easily compiled onto almost any processing architecture, this is not generally true for AQM on programmable data plane hardware, i.e., programmable packet processors. In this work, we highlight corresponding challenges, demonstrate how to tackle them, and provide techniques enabling the implementation of such AQM algorithms on different high speed P4-programmable data plane hardware targets. In addition, we provide measurement results created on different P4-programmable data plane targets. The resulting latency measurements reveal the feasibility and the constraints to be considered to perform Active Queue Management within these devices. Finally, we release the source code and instructions to reproduce the results in this paper as open source to the research community.

cs.NI↗

arXiv2020

On the Throughput Optimization in Large-Scale Batch-Processing Systems

We analyze a data-processing system with $n$ clients producing jobs which are processed in \textit{batches} by $m$ parallel servers; the system throughput critically depends on the batch size and a corresponding sub-additive speedup function. In practice, throughput optimization relies on numerical searches for the optimal batch size, a process that can take up to multiple days in existing commercial systems. In this paper, we model the system in terms of a closed queueing network; a standard Markovian analysis yields the optimal throughput in $ω\left(n^4\right)$ time. Our main contribution is a mean-field model of the system for the regime where the system size is large. We show that the mean-field model has a unique, globally attractive stationary point which can be found in closed form and which characterizes the asymptotic throughput of the system as a function of the batch size. Using this expression we find the \textit{asymptotically} optimal throughput in $O(1)$ time. Numerical settings from a large commercial system reveal that this asymptotic optimum is accurate in practical finite regimes.

cs.PF↗