Jekyll2026-08-31T02:27:54+00:00/feed.xmlYour Site Description Your Name2026-08-30T00:00:00+00:002026-08-30T00:00:00+00:00/mintYour Name

"Abstract", "image"=>"/assets/images/pubpic/mint_cover.png", "text"=>"Generative AI (GenAI) is becoming a mainstream network workload, yet packet-level simulators lack measurement-driven GenAI traffic models. Currently researchers must approximate GenAI services using traditional sources such as file transfer and video streaming, limiting realistic network evaluation of scheduling and capacity planning. We present MINT, a measurement and modeling framework for GenAI network traffic. Using an isolated network-namespace capture pipeline, we collect client-side traces from three LLM providers across four modalities, cloud and edge servers, and wired and wireless network access points. We find that GenAI modalities exhibit distinct upload/download asymmetry and burst structures that differ from traditional applications. MINT clusters and models these burst regimes and reproduces empirical behavior in ns-3 with normalized Wasserstein distances of 2–25%. Our results also reveal that constant token generator models fail to capture realistic packet burst variability. MINT open-sources the first measurement-driven GenAI traffic model for packet-level network simulation."}]]]>

2026-02-12T00:00:00+00:002026-02-12T00:00:00+00:00/Tiny-TwinYour Name

"NextG cellular networks continue to grow in architectural complexity, particularly with the rise of flexible functional splits—techniques that partition the protocol stack between centralized units (CUs), distributed units (DUs), and remote radio units (RUs). These systems span heterogeneous deployments including millimeter-wave, sub-6 GHz, and satellite links, each with distinct characteristics and performance requirements. Developing, testing, and debugging such intricate systems is challenging and requires high-fidelity digital twins that can accurately emulate network behavior at scale. However, existing simulation solutions have fundamental limitations: many rely on specialized hardware accelerators that are expensive and unavailable to most researchers, while others provide only limited fidelity or cannot model the full protocol stack. This gap between simulation accessibility and fidelity has become a significant bottleneck for nextG network innovation and testing.", "image"=>"/assets/images/tinytwin/fig3.png", "image_width"=>800}, {"text"=>"Tiny-Twin addresses these challenges head-on by providing a CPU-native, full-stack digital twin platform purpose-built for NextG cellular networks. Our key innovation is enabling accurate, full-stack simulation of complete network protocols—from physical layer through MAC and RLC to higher layers—while running efficiently on standard computing hardware without specialized accelerators. Tiny-Twin supports multiple functional splits (including split-7.2 and split-8), diverse network topologies, and heterogeneous deployment scenarios. The system is architected to handle realistic traffic patterns, mobility events, and channel conditions while maintaining simulation fidelity. Our extensive evaluation on a real-world O-RAN compliant testbed demonstrates that Tiny-Twin achieves near real-time simulation performance across diverse scenarios—including multi-user environments, high-speed mobility, congestion conditions, and degraded signal scenarios—while accurately predicting network behavior. This makes Tiny-Twin an accessible, practical tool for researchers, developers, and operators to rapidly prototype, validate, and optimize algorithms before deployment to live networks."}]]]>

2026-01-22T00:00:00+00:002026-01-22T00:00:00+00:00/arraylinkYour Name

"The rapid growth of low Earth orbit (LEO) satellite constellations has revolutionized broadband access, Earth observation, and direct-to-device connectivity. However, the expansion of ground station infrastructure has not kept pace, creating a critical bottleneck in satellite-to-ground backhaul capacity. Traditional parabolic dish antennas, though effective for geostationary (GEO) satellites, are ill-suited for dense, fast-moving LEO networks due to mechanical steering delays and their inability to track multiple satellites simultaneously. Phased array antennas offer electronically steerable beams and multi-satellite support. However, their integration into ground stations is limited by the high cost, hardware issues, and complexity of achieving sufficient antenna gain.", "image"=>"/assets/images/arraylink/arraylink_motivation.png", "image_width"=>800}, {"text"=>"We introduce ArrayLink, a distributed phased array architecture that coherently combines multiple small, commercially available panels to achieve high-gain beamforming and unlock line-of-sight MIMO spatial multiplexing with minimal additional capital expenditure. By spacing 16 32x32 panels across a kilometer‐scale aperture, ArrayLink enters the radiative near-field, focusing energy in both angle and range while supporting up to four simultaneous spatial streams on a single feeder link. Through rigorous theoretical analysis, detailed 2D beam pattern simulations and real-world hardware experiments, we show that ArrayLink (i) achieves dish-class gain exceeding that of a 1.47 m reflector, (ii) maintains four parallel streams at ranges of hundreds of kilometers (falling to two beyond 2000 km), and (iii) exhibits tight agreement across theory, simulation, and experiment with minimal variance. These findings pave the way for a practical and scalable approach to boosting satellite backhaul capacity."}]]]>

2025-01-16T00:00:00+00:002025-01-16T00:00:00+00:00/PhaseMOYour Name

"The rapid proliferation of devices and increasing data traffic in cellular networks necessitate advanced solutions to meet these escalating demands. Massive MIMO (Multiple Input Multiple Output) technology offers a promising approach, significantly enhancing throughput, coverage, and spatial multiplexing. Despite its advantages, Massive MIMO systems often lack flexible software controls over hardware, limiting their ability to optimize operational expenditure (OpEx) by reducing power consumption while maintaining performance. Current software-controlled methods, such as antenna muting combined with digital beamforming and hybrid beamforming, have notable limitations. Antenna muting struggles to maintain throughput and coverage, while hybrid beamforming faces hardware constraints that restrict scalability and future-proofing. This work presents PhaseMO, a versatile approach that adapts to varying network loads. PhaseMO effectively reduces power consumption in low-load scenarios without sacrificing coverage and overcomes the hardware limitations of hybrid beamforming, offering a scalable and future-proof solution. We will show that PhaseMO can achieve up to 30% improvement in energy efficiency while avoiding about 10% coverage reduction and a 5dB increase in UE transmit power.", "image"=>"/assets/images/PhaseMO_fig.jpg", "image_width"=>800}]]]>

2024-11-18T00:00:00+00:002024-11-18T00:00:00+00:00/ue-powerYour Name

"With 6.5 billion smartphones in use worldwide, each relying on a battery for key subsystems like display, compute, and cellular connectivity, previous studies on power consumption often used invalidated indirect estimates that failed to isolate specific hardware usage. We address this by utilizing Google's On Device Power Measurement (ODPM) tool for precise power measurements of individual components. Our findings indicate that connectivity (Wi-Fi, 4G/5G) and screen display are the primary power consumers, as shown with the Google Pixel 7A. We also confirmed similar power consumption trends using an energy estimation method on the Samsung S23+. Given the prevalence of smartphones, we discuss the challenges and opportunities for optimizing power usage.", "image"=>"/assets/images/barplot_energy.png", "image_width"=>800}]]]>

2024-11-01T00:00:00+00:002024-11-01T00:00:00+00:00/commradYour Name

"Utilizing monostatic Radar and Bi-static Radio as multi-modal sensing", "text"=>"Millimeter-wave (mmWave) technology is pivotal for next-generation wireless networks, enabling high-data-rate and low-latency applications such as autonomous vehicles and XR streaming. However, maintaining directional mmWave links in dynamic mobile environments is challenging due to mobility-induced disruptions and blockage. While effective, the current 5G NR beam training methods incur significant overhead and scalability issues in multi-user scenarios. To address this, we introduce CommRad, a sensing-driven solution incorporating a radar sensor at the base station to track mobile users and maintain directional beams even under blockages. While radar provides high-resolution object tracking, it suffers from a fundamental challenge of lack of context, i.e., it cannot discern which objects in the environment represent active users, reflectors, or blockers. To obtain this contextual awareness, CommRad unites wireless sensing capabilities of bi-static radio communication with the mono-static radar sensor, allowing radios to provide initial context to radar sensors. Subsequently, the radar aids in user tracking and sustains mobile links even in obstructed scenarios, resulting in robust and high-throughput directional connections for all mobile users at all times. We evaluate this collaborative radar-radio framework using a 28 GHz mmWave testbed integrated with a radar sensor in various indoor and outdoor scenarios, demonstrating a 2.5x improvement in median throughput and an 8x improvement in 20th percentile throughput compared to a non-collaborative baseline.", "image"=>"/assets/images/respic/5G/commrad_collab_learning.png", "image_width"=>800}, {"title"=>"Synchronized radar and radio platform", "text"=>"We built a synchronized 28 GHz Radio and 24 Ghz Radar platform controlled via an FPGA for a trigger-based synchronization.", "image"=>"/assets/images/respic/5G/commrad_sync_radar_radio.png", "image_width"=>800}, {"title"=>"This project won the Qualcomm Innovation Fellowship in 2022", "text"=>"Original award winning QIF poster outlining the original idea of CommRad in May 2022.", "image"=>"/assets/images/respic/5G/commrad_poster_qif.png", "image_width"=>800}]]]>

2024-08-01T00:00:00+00:002024-08-01T00:00:00+00:00/mmsubarrayYour Name

"The future of global connectivity depends on the seamless integration of satellite and terrestrial networks. Satellites and ground stations can now relay data between terrestrial base stations and remote devices, breaking coverage barriers and enabling global connectivity. However, current satellite-to-device architectures require ground stations with extensive fiber backhaul across difficult terrain, leading to high deployment costs, scalability challenges, and increased latency. To address this, we propose JointNets: Joint satellite and terrestrial networks by deploying ground stations within terrestrial coverage areas and using high-speed millimeter-wave links for backhaul. This eliminates the need for long optical fibers, simplifies ground station deployment, enhances scalability, and reduces latency.", "image"=>"/assets/images/mmsubarray/proposed_joint_satellite_5G.png", "image_width"=>800}, {"text"=>"A major challenge in enabling JointNets is that ground station to satellite uplink transmissions (27.5 to 30.0 GHz) interfere with 5G links, leading to reduced efficiency or complete link failure. Current approches, such as distance, frequency, and direction separation, often cause spectrum inefficiency and coverage gaps. This paper introduces mmSubArray: An array of sub-band phased arrays using commercial off-the-shelf phased arrays to enable full-spectrum utilization and joint satellite-terrestrial networks. Our mmSubArray approach splits the bandwidth into overlapping and non-overlapping subbands, with different phased arrays beaming each subband at the base station.", "image"=>"/assets/images/mmsubarray/proposed_mmsubarray.png", "image_width"=>800}, {"text"=>"Enabling Backhaul: Using commercial phased arrays, we demonstrate that simultaneous mmWave transmissions in non-overlapping bands do not suffer significant degradation due to interference. mmSubArray employs one or more phased arrays to beam non-overlapping bands toward interfering ground stations. This allows mmSubArray to enable backhaul for ground stations or support users in interfering directions.", "image"=>"/assets/images/mmsubarray/proposed_mmsubarray_backhaul.png", "image_width"=>800}, {"text"=>"Enabling Coexistence: In the overlapping band, we use the other phased arrays to serve users in non-interfering directions. However, simply beaming in other directions is not sufficient to enable coexistence in many scenarios. We may still have side lobes that fall along the interference directions, causing interference issues. Our key idea is to apply nulling along with beaming in other directions; together, these techniques suppress the interference power below the noise floor and enable coexistence in overlapping bands. Through extensive simulations and real-world measurements, we demonstrate the interference challenges and evaluate the efficacy of our approach.", "image"=>"/assets/images/mmsubarray/proposed_mmsubarray_coexistence.png", "image_width"=>800}]]]>

2024-03-01T00:00:00+00:002024-03-01T00:00:00+00:00/beamarmorYour Name

"BeamArmor is implemented with real-time RAN (srsRAN) with a EdgeRIC controller", "text"=>"", "image"=>"/assets/images/respic/5G/beamarmor_poster.png", "image_width"=>900}]]]>

2023-05-24T00:00:00+00:002023-05-24T00:00:00+00:00/mmspoofYour Name

"FMCW radars are integral to automotive driving for robust and weather-resistant sensing of surrounding objects. However, these radars are vulnerable to spoofing attacks that can cause sensor malfunction and potentially lead to accidents. For instance, an attacker driving ahead of a radar-equipped victim’s vehicle can manipulate the radar signals such that the victim radar measures a false distance to the attacker vehicle (as shown in figure). Upon falsely detecting the attacker’s phantom vehicle, the radar may trigger the vehicle to apply a sudden brake, risking the passenger’s life and causing accidents.", "image"=>"/assets/images/mmSpoof/attack_scenario.png", "image_width"=>800}, {"text"=>"Previous attempts at spoofing FMCW radars using an attacker device have not been very effective due to the need for synchronization between the attacker and the victim. We present a novel spoofing mechanism called mmSpoof that does not require synchronization and is resilient to various security features and countermeasures of the victim radar. Our spoofing mechanism uses a “reflect array” based attacker device that reflects the radar signal with appropriate modulation to spoof the victim’s radar. We provide insights and mechanisms to flexibly spoof any distance and velocity on the victim’s radar using a unique frequency shift at the mmSpoof’s reflect array. We design a novel algorithm to estimate this frequency shift without assuming prior information about the victim’s radar. We show the effectiveness of our spoofing using a compact and mobile setup with commercial-off-the-shelf components in realistic automotive driving scenarios with commercial radars.", "image"=>"/assets/images/mmSpoof/attack_demonstration.png", "image_width"=>800}]]]>

2023-01-26T00:00:00+00:002023-01-26T00:00:00+00:00/dpaYour Name

"How multi-beams support multiple users without loosing power per-beam", "text"=>"Modern mmWave systems have limited scalability due to inflexibility in performing frequency multiplexing. All the frequency components in the signal are beamformed to one direction via pencil beams and cannot be streamed to other user directions. We present a new flexible mmWave system called mmFlexible that enables flexible directional frequency multiplexing, where different frequency components of the mmWave signal are beamformed in multiple arbitrary directions with the same pencil beam. Our system makes two key contributions: (1) We propose a novel mmWave front-end architecture called a delay-phased array that uses a variable delay and variable phase element to create the desired frequency-direction response. (2) We propose a novel algorithm called FSDA (Frequency-space to delay-antenna) to estimate delay and phase values for the real-time operation of the delay-phased array. Through evaluations with mmWave channel traces, we show that mmFlexible provides a 60-150% reduction in worst-case latency compared to baselines.", "image"=>"/assets/images/respic/5G/dpa_hotmobile.png", "image_width"=>800}, {"title"=>"Delay-Phased Array Architecture Details", "text"=>"DPA architecture consists of programmable delay and programmable phase element per antenna with a single-RF chain. These elements can be programmed together to create flexible beam responses that are not possible by either of the two elements alone. Our insight is to control two knobs: delays τn and phase Φn to get the desired response.", "image"=>"/assets/images/respic/5G/dpa_architecture.png", "image_width"=>800}, {"title"=>"Best Poster Runner-up Award at Hotmobile'23", "text"=>"", "image"=>"/assets/images/respic/5G/dpa_hotmobile_poster.png", "image_width"=>800}]]]>