Can Flight Data Recorder Memory be Stored on the Cloud? (Computer Project)

Download Project:

Fields with * are mandatory

ABSTRACT:

Flight data recorders (FDRs, or black boxes) generate data that is collected on an embedded memory device. A well-known difficulty with these devices is that the embedded memory device runs out of space.

To avoid getting into this problematic situation, the software of the FDR is designed to operate in a watchful mode, constantly working to minimize the use of memory space; otherwise a larger FDR would be needed.

However, larger FDRs can be a problem because they have very rigorous requirements; thus, enlargement is costly. Outcomes of this research include the recommendation to send FDR data to a remote cloud storage system, so the data memory device will be unbounded.

SNIPPETS:

Enabling Data Transmission from an FDR to a Storage System:

With the aim of transferring an adequate amount of data from an FDR to the cloud, Compression is required. For such purposes, as with the computational codes that run across today’s FDRs, compression techniques cannot be used arbitrarily. Their use must be dynamically configured to match current requirements (i.e., desired transmission rates  and current platform resources, or network bandwidth and CPU load).

The objective of this paper is to guarantee that the rate of compression speed due to available CPU resources and the compression efficiency will create suitable data volumes transmitted over the network at rates that match current available network resources as well as application requirements.

Compression Methods:

Compression techniques reduce data size by applying compression and decompression techniques to data. This section succinctly reviews the techniques employed in this work, in order to show the trade-offs in using these different techniques.

Huffman Compression:

Huffman coding (Huffman, 1952) has been the first practical compression method. Huffman coding is usually not used in a stand-alone mode (Dandekar, 2013); rather, it is used within more complex compression techniques like JPEG (Wallace, 1991; Wiseman, 2014).

Algorithm.

 Algorithm.

Lempel-ZIV Methods:

The conventional dictionary compression technique is Lempel-Ziv coding (Wiseman, 2007b). WINZIP (WinZip,1998) and gzip (Deutsch, 1996), among other common practical compression tools, employ versions of LempelZiv coding.

The Burrows-wheeler Transformation:

The Burrows-Wheeler transformation (Burrows & Wheeler, 1994) is a dictionary compression technique. This technique utilizes repetitions of words’ sequences in order to improve compression. The technique is lossless (i.e., no information is lost in the compression procedure). Burrows Wheeler transformation outperforms Lempel-Ziv coding; therefore, the use of Burrows-Wheeler transformation in a variety of compression utilities is widespread. However, the execution time of Burrows-Wheeler transformation is normally very long.

Method Comparision:

The attributes of these compression techniques have been evaluated; accordingly, the system will be able to choose the most appropriate compression technique for any given attribute of the data, the available communication bandwidth, and available CPU cycles.

Table 1 qualitatively ranks compression techniques, scaled on four levels:

  • Excellent
  • Good
  • Satisfactory
  • Poor
    Attributes of Comparision Techniques.

        Attributes of Comparision Techniques.

    Figure 2. Compression ratios.

         Compression ratios.

Experimental Results:

FDRs generate various information; therefore, the compression techniques used in this paper (i.e., Huffman, Lempel-Ziv, and Burrows-Wheeler) have been tested with multiple datasets, including a text dataset and a binary dataset. The text dataset includes these parameters:

  • Time;
  • Altitude;
  • Vertical acceleration;
  • Heading;
  • Time of each radio transmission either to or from air traffic control;
  • Pitch attitude;
  • Roll attitude;
  • Longitudinal acceleration;
  • Pitch trim position;
  • Control column or pitch control surface position;surface position;
  • Control wheel or lateral control surface position;
  • Rudder pedal or yaw control surface position;
  • Thrust of each engine;
  • Position of each thrust reverser;
  • Trailing edge flap or cockpit flap control position; and
  • Leading edge flap or cockpit flap control position
Method of comparision.

  Method of comparision.

CONCLUSIONS:

Several benefits have been achieved by placing the FDR  in an IaaS cloud:

(1)  Safer FDR memory;
(2)  Extensive enlargement of the FDR memory; and
(3)  Faster transmission of data.

These results are encouraging. The transmission system is able to select the  Compression algorithm intensity and whether to compress or not at real time based on available CPU cycles and the load on the network line, so the cloud can be a replacement for the internal memory device.

Additionally, clouds are crash-proof. The possibility of a traditional FDR being damaged in a crash is low, because FDRs are designed to sustain even very severe crashes;however if an airplane does crash, potential damage to the cloud is absolutely zero.

Source: Bar-Ilan University
Author: Yair Wiseman

Download Project

Download Project:

Fields with * are mandatory