The Public Safety Innovation Accelerator Program grant opportunity (PSIAP FY17) focuses on research and development projects in six key technology areas:
Mission-Critical Voice (§I.A)
Device-to-Device (Off-Network) LTE Voice Communication
Current public safety land mobile radios (based on P25 or TETRA) provide both indirect communication through a network of repeaters and “direct mode” communication between mobile units without relaying. Direct mode operation is a necessary lifeline for situations in which base-station coverage is unavailable. Currently deployed cellular systems, including LTE, only offer communication through a base station and operator network.
Current standardization efforts for LTE include several device-to-device (D2D) communication capabilities: Proximity Services (ProSe), Vehicle-to-vehicle, LTE-M, and NB-IoT. While the physical layer technology is reasonably mature, there is work to be done in test and verification, as well as commercialization, and in fleshing out higher-layer functionality.
Group Push-To-Talk (PTT) is the core function of current public safety communication systems: At any given time, a user can belong to one of more talkgroups, and a speaker can initiate near-instantaneous broadcast/multicast communication to the selected talk group by pressing the PTT button. PTT functionality is well developed in current narrowband voice radios, and some implementations for broadband cellular networks exist. PSCR seeks to encourage the development, adoption, and evaluation of interoperable standards-based implementations of MCPTT over LTE.
Emergency responders have a compelling need to understand the physical environment in which they are working: Where are public safety personnel and equipment? What hazards and resources are present in the area? What entry and exit routes are available? PSCR refers to the collection of technologies and systems that gather, store, disseminate, and act on location and located information as Location Based Services (LBS).
By far, the most pressing need is to be able to locate first responders indoors (a) with sufficient accuracy to reliably know which room on which floor they are in (b) without requiring that the area in which they’re working has been instrumented in advance. That requires both “raw” positioning and reliable and timely dissemination of position information.
Over the next decade, public safety responders will increasingly be connected to mobile broadband networks, enabling a far greater quantity and variety of data to be collected and exchanged. At this point, it’s largely an open question how that data might be used, but some core questions keep coming up: Increasing communication bandwidth doesn’t change human cognitive capacity – how can automation help responders (and commanders, dispatchers, etc.) identify and make use of relevant information without causing distraction or cognitive overload? For example, can visual processing identify dangers (e.g. weapons) or assets (e.g. exit routes, AEDs, fire standpipes) from images and video feeds? Or can speech (and background audio) processing identify especially-critical responder calls and call attention to them? What kind of data is available (or could be made available)? How can data be used responsibly: How can potentially sensitive data be stored, shared, and worked with in ways the preserve confidentiality and privacy? How can untrusted, publicly-sourced data, which is often informative, but also sometimes wrong, and sometimes maliciously so, be incorporated? How can we guard against bad data (or bad analysis) leading to bad decisions? What tools and processes can public safety agencies use, and how are their needs different from scientific or business data analysts?
Because the field of analytics is so large, and public safety’s needs are still largely undefined, PSCR is presently interested in exploratory research involving public safety agencies: To identify and make available existing (or newly-created) data sets, to develop/integrate tools for data handling and analysis, and to identify needs and opportunities for analytics.
PSCR seeks to develop a comprehensive model (or set of models) describing the communication needs associated with public safety operations. The goals are to (1) provide measures against which to evaluate proposed communication technologies, (2) enable planning and provisioning of operational systems, and (3) provide a common reference for modeling specific incident response scenarios. To do this, it is necessary to go beyond aggregate measures of load and also capture some of the semantics and operational significance of communications: For example, a useful model would distinguish between a video feed showing a hostage situation and one monitoring highway traffic volume: They might naturally have the same bitrate, but a reasonable prioritization system would handle them very differently, and degradation or delay of the former is a greater failure than degradation or delay of the latter. Additionally, a model should include enough location information to reason about radio propagation and coverage between communicating parties.
We expect that successful modeling efforts will draw on a variety of data sources, such as: Agencies’ written plans and procedures, after-action reports from incidents, machine-generated logs and recordings from communication systems and dispatch systems, and participant interviews. We expect to support research into communication practices and needs across the spectrum of agencies, operations and incidents. This can include both “raw” data gathering and explicit model building.
Application and Network Simulation and Modeling
We wish to create a standard, shared pool of tools and resources for public safety communication research. Simulation and modeling tools are central to networking research, but many public-safety-relevant technologies are insufficiently supported by popular research tools. Some key technologies are: LTE, specifically with new/proposed features such as Device-to-Device communication, eMBMS & GCSE, IMS/VoLTE, and MC-PTT, and IP-over-LTE breakout and aggregation (e.g. LIPA, SIPTO, IFOM); satellite and microwave backhaul; Land Mobile Radio (e.g. P25); and interconnections between the preceding. In addition, support for indoor mobility and propagation – especially in ways that enable joint simulation of positioning and communication systems – is desirable. We are interested in both radio-level (that is physical/link) and network-level (that is, packet-oriented discrete event) simulation tools where each is appropriate.
Programmable / Software Defined Radios
As with the simulation and modeling above, we are interested in creating LTE prototyping platforms for use by both PSCR and external researchers. The goal is to have programmable/SDR implementations that are sufficiently complete to build and test features which are not yet present in commercially-available devices. These systems need to be sufficiently complete to run a full LTE stack using only open-source components, and sufficiently interoperable to use SDR UEs with commercial eNodeB and core implementations, and vice-versa. The initial priority will be on the UE and eNodeB, but core services and application servers are also of value.
Public safety communication systems must continue to work – and indeed may be most necessary – under circumstances in which other networks fail: Individual responders will need to operate in areas without cellular network coverage. Groups of responders will need to bring coverage with them, in scenarios ranging from a handful of police officers patrolling a remote road to thousands of wildland firefighters and support staff with semi-permanent command posts and camps. Backhaul connections, if they exist at all, will be much more limited in throughput, reliability, and latency than the local networks. Disasters may take out power, backhaul, or cellular towers over significant areas. Jamming (deliberate or accidental) may impair specific frequencies over a wide area. We wish to ensure that – up to the limit of what is physically possible – public safety communication (including access to data and computation) continues to function in such scenarios.
Challenges which we have identified include: Decentralizing existing LTE/IMS functionality; handling IP breakout and IP mobility to ensure efficient and robust routing; data replication, delivery, and consistency in disconnect and partially-connected environments; service (think e.g. cloud/mobile GIS applications) replication, availability, and consistency; practices, abstractions, APIs, etc. to support developing “future-proof” applications that can take advantage of improved host, data, and service mobility; and security (identity and access control) approaches that are robust to disconnected operation, including practical “in-the-field” management.