Live streaming services power mainstream real-time interactions on top of dedicated live streaming networks (LiveNets). Yet making LiveNets reliable at scale is challenging: failures arise on the userfacing delivery path and within streaming protocol and application logic, so operators need both continuous runtime monitoring to detect and localize incidents quickly and proactive preflight testing to exercise changes under representative environments and sustained playback behavior. Meeting these goals hinges on the right vantage point: the observability workflow must traverse the same network paths and delivery stacks as users while remaining controllable and non-intrusive. We present Horizon, which leverages near-user, provider-managed hyper-edge devices and orchestrates them into a shared fleet that supports both always-on monitoring and customizable, scenario-driven validation. Horizon has been deployed in production for over three years; in 2025, it identified 2,000+ major network incidents using 100,000+ hyper-edge agents.

We propose remote direct code execution (RDX), which elevates the power of RDMA from memory access to code execution. We target runtime extension frameworks such as Wasm filters, BPF programs, and UDF functions, where RDX enables an agentless architecture that unlocks capabilities such as fast extension injection, update consistency guarantees, and minimal resource contention. We outline the roadmap for RDX around a new CodeFlow abstraction, encompassing programming remote extensions, exposing management stubs, remotely validating and JIT compiling code, seamlessly linking code to local context, managing remote extension state, and synchronizing code to targets. The case studies and initial results demonstrate the feasibility of RDX and its potential to spark the next wave of RDMA innovations.

The interference graph, depicting the intra- and inter-cell interference channel gains, is indispensable for resource allocation in multi-cell networks. However, there lacks viable methods of interference graph estimation (IGE) for multi-cell multi-numerology (MN) networks. To fill this gap, we propose an efficient power-domain approach to IGE for the resource allocation in multi-cell MN networks. Unlike traditional reference signal-based approaches that consume frequency-time resources, our approach uses power as a new dimension for the estimation of channel gains. By carefully controlling the transmit powers of base stations, our approach is capable of estimating both intra- and inter-cell interference channel gains. As a power-domain approach, it can be seamlessly integrated with the resource allocation such that IGE and resource allocation can be conducted simultaneously using the same frequency-time resources. We derive the necessary conditions for the power-domain IGE and design a practical power control scheme. We formulate a multi-objective joint optimization problem of IGE and resource allocation, propose iterative solutions with proven convergence, and analyze the computational complexity. Our simulation results show that power-domain IGE can accurately estimate strong interference channel gains with low power overhead and is robust to carrier frequency and timing offsets.

Resource allocation in multi-cell networks typically requires knowing the inter-cell interference channel gains. When multiple numerologies are employed, resource allocation further needs to estimate the inter-numerology interference within each cell. However, there lacks viable methods of interference graph estimation (IGE), depicting the intra- and inter-cell interference channel gains, for multi-cell multi-numerology networks. To fill this gap, we propose an efficient power-domain approach to IGE for the resource allocation in multi-cell multi-numerology networks. Unlike traditional reference signal-based approaches that consume frequency-time resources, our approach uses power as a new dimension for the estimation of channel gains. By carefully controlling the transmit powers of base stations, our approach is capable of estimating both the intra- and inter-cell interference channel gains, useful to resource allocation. Further, as a power-domain approach, it can be seamlessly integrated with the resource allocation in multi-numerology networks, such that IGE and resource allocation can be conducted simultaneously. We derive the necessary conditions for power-domain IGE and formulate the joint optimization problem of IGE and resource allocation. Our simulation results show that power-domain IGE can accurately estimate the interference channel gains and incurs low power overhead.

Traditional wisdom for network resource management allocates separate frequency-time resources for measurement and data transmission tasks. As a result, the two types of tasks have to compete for resources, and a heavy measurement task inevitably reduces available resources for data transmission. This prevents interference graph estimation (IGE), a heavy yet important measurement task, from being widely used in practice. To resolve this issue, we propose to use power as a new dimension for interference measurement in full-duplex millimeter-wave backhaul networks, such that data transmission and measurement can be done simultaneously using the same frequency-time resources. Our core insight is to consider the mmWave network as a linear system, where the received power of a node is a linear combination of the channel gains. By controlling the powers of transmitters, we can find unique solutions for the channel gains of interference links and use them to estimate the interference. To accomplish resource allocation and IGE simultaneously, we jointly optimize resource allocation and IGE with power control. Extensive simulations show that significant links in the interference graph can be accurately estimated with minimal extra power consumption, independent of the time and carrier frequency offsets between nodes.

Traditional wisdom for network resource manage-ment is to allocate separate frequency-time resources for measurement and data transmission tasks. As a result, the two types of tasks have to compete for resources, and a heavy measurement task inevitably reduces available resources for data transmission. This prevents interference graph estimation (IGE), a heavy yet important measurement task, from being widely used in practice. To resolve this issue, we propose to use power as a new dimension for interference measurement in full-duplex mmWave backhaul networks, such that no extra frequency-time resources are needed for measurement. Our core insight is to consider the mm Wave network as a linear system, where the received powers of a node can be expressed as the product of the powers of transmitters and the equivalent channel gains from the transmitters to the node. By controlling the powers of transmitters, we can find unique solutions for the equivalent channel gains, which will then be used to estimate interference. To accomplish resource allocation and IGE simultaneously, we jointly optimize resource allocation and IGE with power control. Extensive simulations show that significant links in the interference graph can be accurately estimated with less than 3 % increase in power consumption, independent of the time synchronization and carrier frequency offset (CFO) estimation errors between nodes.