Department 13 Patents & Advanced Technology
Department 13’s patented technologies take advantage of the basic principles of radio wave physics to identify signals, control radio systems, and enhance data bandwidth. At Department 13, whether we are implementing new tactical systems or next-generation communication and automation networks, our solutions provide the best performance, highest efficiency, and powerful results.
Airborne Relays SA-004
- 3 issued US patents, 1 pending US application
- International: AU, CN, EP, IL, JP
Covers navigation and distributed processing by a swarm of autonomous agents (sUAS’s, robots, etc). Swarm intelligence refers to the study of the collective behavior of multi-component systems that coordinate using decentralised controls and self-organization.
From an engineering point of view, swarm intelligence emphasises the bottom-up design of autonomous distributed systems that can provide adaptive, robust, and scalable behaviors.
Swarm intelligence includes large networked systems such as sensor and actuator monitoring networks, supply-chain management systems, the power grid, and traffic systems. An important aspect of swarm intelligence is the command and control of teams of sUAV’s, such as surveillance and combat scenarios.
The focus on navigation decision-making based on sensor readings onboard each autonomous agent is evident. Onboard sensors can include radio receivers, radar, lidar, acoustic sensors, cameras and imaging systems. When an agent receives a sensor reading, it collaborates with other agents to analyse the source of those readings, which can include adapting its navigation to obtain more readings. The agents or a centralised system can perform threat analysis and may perform counter-measures, including mitigation.
SA-001, SA-002: Coordinated Multipoint, & RLNC
- SA-001, SA-002: 2 issued patents, 1 pending
- RLNC: 5 issued patents, 1 pending
Geographically distributed receivers collect transmissions from multiple sources, and these signals are collected in a central location where they are then separated. Yes, this is extremely broad, but it claims priority to 2001.
This is useful in a DIAL system that uses multiple radio receivers, and it is useful for low altitude airspace situational awareness where many targets need to be tracked. The number of transmitting sources the system can differentiate is equal to the number of radio receivers. If the system is unable to differentiate all of the transmit sources, it knows to increase the number of radio resources until their number is sufficient to separate the signals. The claims in the RLNC patents are also quite broad, and they cover this with language that says, “selecting a sufficient number of receive nodes to enable the destination node to resolve the original data signals”.
SA-003: Distributed SDR
- 1 issued patents, 2 pending
In this patent, radio processing is split up between two devices connected via a network. When a radio signal is received at a receiver, some radio processing (such as filtering and down-converting the signal to baseband) is performed at the receiver before it is sent to another radio processor, which performs additional radio processing (such as demodulation and decoding). MESMER does this because the raw data bandwidth from the antenna to the computer would otherwise be too high. Almost all DIAL systems do this.
- 1 issued patent, 1 pending
This is a Device-to-Device (D2D) communication link in the 5G standard that would likely be used to control consumer drones. In this case, a cell tower allocates a channel to a controller or smartphone, which then uses the channel to communicate with a drone. The tower knows that the channel is intended for a drone, so it doesn’t need to listen to the transmissions. This patent family is similar to the Spectrum Sharing patent family (below). A DIAL system will use this to detect D2D drone-controller links.
Spectrum Sharing (D2D)
- 4 issued patents, 1 pending
This D2D format is being considered for the 5G standard. Unlike centralised D2D (in CP above), decentralised scheduling of D2D links is permitted. The advantage is that it eliminates the signaling overhead and delay of centralised D2D control, which are recognised problems in 5G. The low-latency benefit of this form of D2D makes it attractive for collision avoidance in autonomous navigation, which enables drones to quickly resolve conflicts. Like commercial air traffic control, UTM will likely include a system like TCAS. In commercial applications, drones can communicate with ground-based sensors that have limited range due to power constraints and terrestrial channel impairments, which would be useful across a broad range of industries. One application is 5G smart warehouses for inventory management and environment sensing.
007: OFDMA Power Efficiency
- 3 issued patents, 4 pending
In the downlink (from the tower to the phone), each user signal is in a different frequency channel, which means SC-FDMA won’t work. This normally isn’t a problem because cell towers aren’t battery-powered, so power efficiency isn’t as big of a concern. This isn’t the case in 5G because cell towers are being replaced with low-power micro-cells, and smartphones can relay signals to other smartphones. Also, drones are being used as relays in 5G and ad-hoc networks.
Few people understand that the code space in SC-FDMA can be multiplexed such that different SC-FDMA codes carry different data streams. This allows all the user signals to be in the same frequency channel, so SC-FDMA improves power efficiency. This is ideal for control signaling in UTM and commercial drone services whereby a single low-power signal has different channels in order to control multiple drones. Similarly, a drone can transmit different data channels inside the same low-power signal. The applications span a wide range of commercial drone services and apply generally to when a drone has multiple sensors or data feeds or relays signals from other drones or ground-based sensors.
CI-010: Blind Adaptive Receiver
- 1 allowed patent, 3 pending
Spectrally decomposes a received signal and blindly decodes its frequency components to estimate the transmitted data. This covers the DIAL part of MESMER and Animal and is pertinent to almost all DIAL systems.
The patent strategy here is clever. The nearest relevant art is from 2009, but the priority date for this patent family is 2002. The 2002 disclosure describes air-traffic control communications, which would allow us to broadly cover receiving any control communications for aircraft. However, the 2002 case was abandoned due to a postal delay in filing the issue fee, and the continuing case (C-MIMO) filed in 2005 does not disclose air-traffic control communications. So, we filed a petition to revive the 2002 case and filed continuing applications directly from it that include the required disclosure. Furthermore, the 2005 case supports our broader blind-adaptive receiver claims, so we might be able to claim priority to that case instead of the 2002 case to get an extra three years of the patent term.
- 5 issued patents, 2 pending
This contains broad disclosures and claims that are applicable to all types of communications including wireless sensors, vehicle communications and RFID systems. Generally, it describes how to combine antennas residing on different platforms and select which antennas to use in order to get the best signal. This is especially useful for separating interfering signals, which is why it is MIMO and why it is used in current cellular networks. The patent disclosure describes how to adapt the algorithms to account for the antennas being far apart. Next, it goes into a lot of detail about how a received signal can be decomposed into its frequency components, and then blind source separation algorithms can separate or detect interfering signals. This is analogous to separating signals received by multiple antennas, but it performs the same function based on different frequencies instead. Previous blind source separation was based on information about the channel, which tells you how signals are mixed. The novel aspect here is that it uses information about the signals (such as modulation, bandwidth, spreading, etc), which it might learn or simply guess in order to separate the signals. Today, this approach is widely used in signal analysis particularly in DIAL. These cases give us priority dates of 2002 and 2004.
- 5 pending patents
Artificial Neural Networks (ANNs) learn by employing back-propagation which is time-consuming and requires substantial computing resources. Instead, ANNIE models the ANN as a linear network, and its nonlinear activations are just linear updates to the model. This can be thought of as digitising an analog function, and this digital version of the ANN is a good approximation of the actual ANN. This allows a set of clever linear algebra tricks to replace backpropagation, which might speed up learning by orders of magnitude. The nature of the linear model can reflect the type of solution to be learned, which means that one linear model is optimised for learning new drone control signals while another linear model is optimised for learning new video downlink signals. ANNIE’s learning is so much more efficient it allows a “teacher” AI to teach student ANNs how to select their linear models. It can enable many nested layers of teacher-student relationships.
This is useful for Animal but can also be useful for a wide range of airspace management, surveillance, data analytics, autonomous navigation, and business optimisation applications.
- 5 issued patents, 1 pending
The issued patents cover Content Delivery Networks, which are not useful to Department 13. In reviewing these cases, some useful disclosures were noted relating to D2D. The pending case has D2D claims and has received a non-final office action which will be useful for Department 13.
Radios used to control drones employ well-defined techniques to begin communications and maintain a radio link. By adapting to the protocol used to control a drone, MESMER inserts messages that tell the drone to exit restricted airspace, return home, or land in a predetermined safe zone. MESMER can simultaneously control drone swarms (multiple drones) that use different radio protocols.
Drone Identification and Classification
MESMER uses multiple layers of blind adaptive analysis to identify a drone’s type and radio protocol. When MESMER detects a previously unidentified drone, it identifies the drone’s modulation type, format, and other signal information that may be used to redirect the drone.
Distributed Software Defined Radio (SDR)
Detecting and communicating with a large number of drones can quickly overload the processing capability of a centralised system. But when different radio processing operations are implemented in software and spread across a network, a system can dynamically add more sensors and processors to adapt to a changing threat scenario. A virtual SDR can be assigned to each targeted drone and then migrate between processors and sensors to follow the drone as it moves through a network.
Tactical Network Dominance
Tactical Network Dominance is a technique that begins with blind signal analysis. The analysis characterises signals, identifies radio networks, and performs a threat assessment. A tactical response includes protocol manipulations that exploit vulnerabilities in an enemy’s radio network. Messages may be intercepted, service may be denied, and the network may be penetrated as part of a broader system infiltration strategy.
The distributed processing techniques in Tactical Network Dominance are adapted to protect the network infrastructure and users from a variety of attacks. Radio behavior analysis and RF fingerprint tracking are techniques that a hostile entity might utilise to track and identify users. Blue Sphere helps users anonymize these signatures to reduce the threat of an attack by an outside entity. The network infrastructure is protected from infiltration and denial-of-service attacks using a distributed monitoring and authentication system that is modeled after the human body’s immunological functions.
Low Probability of Interception / Low Probability of detection (LPI/LPD) protects communication links from eavesdropping. MESMER’s approach breaks all the rules by using other peoples’ communication signals to carry our information. The information is encoded in signal distortions that are normally filtered out of received signals. Advanced forms of this technique are implemented in MIMO channels.
Radio Resource Sharing and Parasitic Networks
The radio spectrum is a scarce resource, particularly in the lower frequency bands which have the best propagation characteristics. A user in a first network can repurpose scheduled radio channels to avoid interfering with the first network. When a downlink channel is assigned to a cellular user, only that user device is listening on that channel. This enables the user to concurrently employ the assigned channel for other communication links. Uplink channels are repurposed by embedding a second network’s transmission protocol in the data payload of the first network’s transmission frames.
Cooperative- Multiple Input Multiple Output (C-MIMO)
Cooperative- Multiple Input Multiple Output (C-MIMO) is also known as Virtual MIMO, Distributed MIMO, and Network MIMO. C-MIMO solves some of the most important problems in radio communications and dramatically reduces the network operator’s OpEx and CapEx. There are two basic forms of C-MIMO: server-side and client-side.
Server-side C-MIMO is employed in a distributed antenna system, such as multiple cellular base stations that are connected to a central processor, or cloud. The processing exploits the natural scattering environment to function as a giant lens. Buildings, hills, and trees focus radio transmissions to produce a tiny coherence zone at each client device. This allows the network to serve each client with the full spectrum, instead of dividing the spectrum (and thus the data bandwidth) between users who join the network.
Client-side C-MIMO can work in tandem with server-side C-MIMO to enable client devices to share each other’s coherence zones. The sharing results in the unusual condition that as more users join the network, the available data bandwidth increases rather than decreases. Also, since users become part of the network infrastructure, coverage and network capacity can dynamically adapt to demands without costly capital expenditures.
Cooperative-MIMO is expected to be the operating framework for next-generation counter-sUAS because it is a cloud-based computing system that operates with a distributed sensor network. This enables novel capabilities that enhance counter-sUAS, such as highly localised electronic countermeasures against targets, new artificial intelligence paradigms, and effective responses to sUAS swarms.
Cooperative Subspace Coding (CSC)
Cooperative Subspace Coding (CSC) is likely the most efficient form of Linear Network Coding. It improves network efficiency for all data communications, including terrestrial wireless networks, satellite networks, and wired networks. Published results show a 5 to 20-fold increase in data bandwidth. CSC also facilitates cloud storage and channel bonding and enables highly efficient file-sharing networks with less management overhead. CSC benefits the counter-sUAS market by enhancing the bandwidth and reliability of both front-haul and back-haul networks that are necessary for communicating large volumes of data in counter-sUAS sensor networks.
Single Carrier FDMA (SC-FDMA)
Using a single line of code in our software-defined radio, Department 13 reshapes transmitted signals to reduce their dynamic range. This results in many benefits, including longer battery life, less-expensive power amplifiers, reduced interference, and improved performance. Single Carrier FDMA is important for drone communications and any other battery-powered wireless devices.
Coordinated Multipoint is an essential part of the LTE-Advanced cellular specification. Coordinated Multipoint aspects of Cooperative-MIMO will be essential to all distributed networks, especially counter-sUAS networks.
Cloud Radio Access Network (C-RAN)
Cloud Radio Access Network (C-RAN) is one of the most important elements of 5G network architecture. C-RAN greatly reduces the network operator’s OpEx and CapEx. C-RAN provides an efficient framework for artificial intelligence in counter-sUAS by efficiently partitioning signal processing operations between the network edge and the core network. This optimises processing and bandwidth resources.
Airborne Relays is an innovation in cooperative radio networks that uses highly mobile airborne platforms as a distributed sensor network. This is essential for drone air-traffic control, next-generation cellular networks, and counter-sUAS. Like a swarm of birds, each sUAS employs an autonomous navigation system that uses a simple rule base to avoid collisions and to optimise flight patterns. This rule base is adapted to optimise radio channels, detect target sUAV’s, and perform electronic countermeasures. Airborne relays may be quickly deployed. When used to provide wireless network services, the cost of CapEx and OpEx are a fraction of traditional networks.
For more information please contact Jonathan Hunter.