Skip to content
Tekt Industries

Specialisation · Design · Manufacture · Lifecycle

DSP Engineering

Signals into decisions. We design and implement DSP pipelines — on FPGA, microcontrollers and edge silicon — for radar, audio, biosignals and RF.

Discuss a DSP project

Why Tekt

Algorithms tuned to the silicon they run on

DSP value is won in implementation: an algorithm that is elegant in simulation but blows the power budget on target hardware is a failed algorithm. Our engineers take pipelines from mathematical design through fixed-point optimisation to verified implementation on FPGA and microcontroller targets.

From bionic vision processing to passive radar systems, our DSP work runs in some of the most demanding signal environments in Australian industry.

FPGA signal chainsFixed-point & numerical optimisationRadar & RF signal processingAudio & biosignal pipelinesDetection, tracking & classification

How we deliver

Unique value, end to end

  • Model to metal

    Reference models proven in simulation, then implemented cycle-accurate on target.

  • Power-budget honest

    Every pipeline benchmarked for MIPS, memory and milliwatts on real hardware.

  • Verified against truth

    Captured real-world datasets validate performance before and after deployment.

Case studies

Case studies

Real programs from this practice — the brief, the engineering approach, and where it landed.

  • Case study 01 · Silentium Defence

    Passive radar processing chain

    Detection without transmission demands extraordinary signal discipline. We implemented clutter suppression and tracking pipelines for Silentium’s passive radar systems, proven against real captured environments.

  • Case study 02 · Bionic Vision Technologies

    Implant signal processing

    Translating camera input into stimulation patterns in real time — we engineered the DSP chain at the heart of the bionic vision processing system under medical-grade verification.

  • Case study 03 · Nura

    Adaptive audio pipelines

    Nura’s hearing-adaptive audio demanded psychoacoustic processing on consumer silicon. We helped mature the DSP foundations that made personalised sound a shipping product.

What we solve

Problems this practice removes

  • Algorithms elegant in simulation, impossible on target
  • Noise floors that bury the signal of interest
  • Latency budgets blown by naive pipelines
  • No ground truth to prove performance

Tools, processes & standards

MATLAB/Python reference modellingFixed-point & CMSIS-DSP optimisationFPGA toolchainsCaptured-dataset validation rigsSpectrum & vector analysis labRequirements-linked verification

Who delivers it

The team behind it

  • DSP engineer — Deep signal theory with FPGA and MCU delivery
  • RF systems engineer — Radar and RF chain design and measurement
  • Verification engineer — Dataset curation and metric-driven proof

Capabilities

Filtering & detectionFPGA pipelinesFixed-pointValidation

FAQ

Frequently asked questions

Have a signal problem?

We turn noisy real-world signals into product-defining features.

↑ Back to top
Contact us