Wide-Area Infrared Sensor for Persistent Surveillance (WISP)

The objective of the Wide-Area Infrared Sensor for Persistent Surveillance (WISP) program is to improve nighttime aerial persistent surveillance by using readily available detector array technology mated to a readout integrated circuit (IC) designed to implement the digital-pixel focal plane array concept. WISP produces a 15-km-diameter circle of persistence (CoP) on the ground from an 18-kft altitude with 0.8-m average ground sample distance (GSD).

Airborne WISPFigure 1. High-resolution WISP image. (Click on image to view larger version and detailed caption.)

Digital-Pixel Focal Plane Arrays (DFPAs), developed at Lincoln Laboratory, are a key enabling technology for WISP. DFPAs allow a very large scene to be scanned quickly without stopping and stabilizing the system to acquire individual frame images. The WISP sensor consists of a modestly sized DFPA coupled to a high-speed, continuous-motion scanner. The DFPA and scanner are mounted to a three-axis continuously rotating gimbal. WISP acquires multiple wide swaths of image data that are stitched together in software to render a large (850 Mpixel) two-dimensional image frame.

The current WISP sensor utilizes a 640 x 480 scanned sensor array. The current version, although still being used in experiments, has the capability for useful deployment.

Fast and continuous-motion scanning over a wide field of regard (FOR) is enabled by the unique DFPA readout architecture. Figure 2 shows the general structure of the DFPA architecture schematically. The architecture for  specific devices varies depending on requirements.  Photocurrent is integrated onto a small capacitor through an amplifier. When the voltage on the capacitor reaches the threshold voltage of a comparator, a pulse is created that drives the input of a sequential counter. At the same time, the capacitor is reset. The reset operation is very short, on the order of a picosecond or less. The capacitor, comparator, pulse generator, and reset logic form a current controlled oscillator (CCO). The output frequency of the CCO is directly proportional to the current produced in the pixel photodiode. The counter acts as a digital integrator, incrementing on every CCO pulse.

DFPA block diagramFigure 2. DFPA block diagram showing (a) the unit cell architecture and (b) the orthogonal transfer structure. Unit cell is 30 microns square for the existing 640 x 480 DFPA. (Click on image to see larger version.)


Following an integration period, the CCO is disabled. Each counter then acts as a static memory register in each pixel to store the digital value. Each memory register is configured as a parallel shift register and connected to each of its four orthogonal pixel neighbors; i.e., upon command, digital pixel data can be transferred to any one of four neighboring pixels. The operation is global so that the entire array is transferred upon command. Since the operation is also parallel, the image array can be transferred one column (or row) in a single clock cycle. Structures on the periphery of all four sides of the pixel array receive the parallel data and serialize it for transfer off the DFPA though output amplifiers. To read out the array, digital data are quickly clocked to transfer column (or row) data repetitively in one direction until all data have been transferred to the periphery output registers and then read out though input/output (I/O) structures. The shift operation is completely controllable; the readout direction can be altered on the fly, shifting can be stopped at any time, and the counters can be re-enabled to count.

WISP TDI scan schematic diagramFigure 3. (a) WISP time-delay and integration (TDI) scan schematic diagram top view. A fast spinning polygon is used to reflect the field of view; as the polygon scans x deg, the FOR is scanned 2x deg. (b) Side view of the sensor mounted in a gimbal and rotated through 360 deg to generate the second scan dimension. (Click on image to see larger version.)


WISP utilizes the DFPA architecture to accomplish digital time-delay and integration (TDI). Figure 3 shows the WISP concept schematically. A fast spinning polygon provides the cross scan for TDI operation. The polygon can spin over 1000 rpm, presenting a fast, continuously scanned image to the DFPA. To accomplish TDI, the DPFA follows the following steps:

  1. Integrate for approximately 0.5 microseconds
  2. Disengage the counters
  3. Shift the array one column
  4. Re-engage the counters
  5. Repeat

The shift operation takes on the order of 100 nanoseconds so that there are few lost photons.

Illustration of WISP TDI scanningFigure 4. Illustration of WISP TDI scanning. Polygon scans DFPA FOV along a ground swath. Gimbal rotates so that next swath is adjacent with some overlap. (Click on image to see larger version.)

The schematic in Figure 4 shows conceptually how the data are collected on a two-dimensional scene. The WISP scan FOR is approximately 150° x 100°.  Image resolution (or ground sample distance) is determined from the characteristics of the optical system and DFPA.

One benefit of the TDI scanning operation is that each image pixel is the average of N samples, where N is the number of columns in the DFPA, which mitigates the effect of a few bad pixels and leads to 100% effective operability when the DFPA is operating in TDI mode.


top of page