The project Large Object Transmission Under Safety Constraints (LOTUS) addresses challenges in the communication of mobile systems, like Vehicle-to-Everything (V2X) or mobile robots.
Numerous roadmaps and surveys within the scope of Vehicle-to-Everything communication envision an increasing amount of demanding services. Examples are cooperative perception and remote driving services, which are crucial to further enhance automated mobility (SAE Level 4+). These applications require the reliable exchange of large data objects in real-time, such as raw or preprocessed sensor data.
Wireless communication is inherently lossy due to fading effects like reflections, shadowing, and Doppler shift, impacting the signal-to-noise ratio (SNR) and causing bit errors. Additionally, vehicle mobility necessitates constant roaming - and thus frequent handovers - between access points (APs), disrupting data transmission and compromising reliability. Resilient and continuous application operation requires robustness against these failure modes.
Current Vehicle-to-Everything (V2X) standardization efforts mainly involve the exchange of small data for sharing status and intent information. Hence, reliability and low-latency mechanisms are optimized for small, single-packet data. Such mechanisms are not able to consider a packet as part of a data object’s context, and thus insufficiently address the constraints of large data object transmissions. Consequently, novel solutions for a reliable transmission of large data objects under real-time constraints are needed.
For this purpose, we propose application-centric solutions that overcome the issues of state-of-the-art V2X/wireless technologies. Specifically, application-centric approaches allow for utilizing application-information/knowledge for sophisticated coordination, exploiting sample-level deadlines and inherent slack for robustness against various error sources.
This website provides a brief summary of ongoing work and references to publications. Additionally, current test setups and recent measurement and simulation data are reported.
Recent Activities
Check out youtube for our video demonstrating the low-latency continuos connectivity and streaming approach using a physical demonstrator setup. Using a combination of W2RP and low overhead link monitoring we manage to achieve deterministic handovers in less than 10ms, thereby enabling reliable streaming of large data despite the challenges caused by roaming situations. More information
We published our opensource lighweight W2RP (lwW2RP) implementation on github.
We published a short overview paper on the current status of the LOTUS project. The paper can found here.
Our work on improving sample transmission robustness by exploiting data dynamics prevalent in (cooperative) perception applications has been published in the IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems (TCAD) and presented at ACM SIGBED International Conference on Embedded Software (EMSOFT). More information
Jonas Peeck successfully defended his PhD thesis titled “Efficient Real-Time Communication and Processing of Large Data Objects in Autonomous Systems”, which includes discussions on the wireless exchange of large objects. The thesis can be found here.
Today marks the official start of the LOTUS (Large object Transmissions under Safety Constraints) project, funded by Deutsche Forschungsgemeinschaft DFG (German Research Foundation).
While our previous works on application-centric resource management addressed the coordination of multiple applications and network segments, they so far lacked detailed descriptions on how safety can actually be achieved given such a resource management approach. Our latest IEEE Transactions on Vehicular Technolgy (TVT) article addresses these limitations by proposing a safe and synchronized reconfiguration protocol. More information
At IEEE Intelligent Vehicles Symposium (IV) we presented both our works on Shared Slack Budgeting for Ultra Reliable Hard Real-Time Streaming (paper) and our first work on enabling continuous connectivity and large data streaming in roaming scenarios (paper). More information: Shared Slack and Continuous Connectivity.
Reliable multicast of large samples is possible using WiMEP. The corresponding article can be found in the ACM Transactions on Cyber-Physical Systems (TCPS). More information
At IEEE IECON 2023, we presented our proposal on overlapping sample transmissions. This W2RP extensions has been designed in order to, among others, better cope with more challenging burst scenarios. More information
An extension of W2RP designed for coordinating multiple applications has been presented at the Conference on Real-Time Networks and Systems (RTNS). We proposed a decentralized and dynamic adaption mechanism that proved highly effective in managing multiple large data application. More information
We extended our initial work on W2RP by presented formal proofs of reliability and showed that given awareness of application properties, reliability guarantees can be given that ourperform existing solutions. The article has been published in the IEEE Transaction on Vehicular Technology (TVT). More information
We released our OMNeT++-based IDA Wireless Simulator on Github. It supports both W2RP and the multicast WiMEP.
We proposed a centralized application-centric Resource Management (RM) approach that combines and coordinates application properties and demands as well as network layer constraints to address the coordination of various wireless applications across different network segments. The article was published as part of the ACM Transactions on Embedded Computing Systems (TECS). More information
Our first paper on enabling reliable sample transmissions has been published and presented at the IEEE Real-Time Systems Symposium (RTSS) 2021. The paper introduces the Wireless Reliable Real-Time Protocol (W2RP) and showcases its improvements over the MAC layer retransmissions of 802.11 as well as the off-the-shelf DDS middleware. More information