DOCSLIB.ORG

Data Compression Techniques: a Comparative Study

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Data Compression Techniques: a Comparative Study International Journal of Applied Research & Studies ISSN 2278 – 9480 Review Paper Data Compression Techniques: A Comparative Study Authors: 1Pooja Jain*, 2 Zeeshan Khan, 3Anurag Jain Address For correspondence: 1, 2, 3 Dept. of Computer Science, Radharaman, Institute of Technology and Science, Bhopal, India. Abstract- Data compression is one of many technologies that enables and images but have made compression of such signals central today’s information revolution. Image compression is an application of to storage and communication technology. data compression. Image compression is now essential for applications such as transmission and storage in data bases. In this paper we review II. Principles Behind Compression and discuss about the image compression, need of compression, its A common characteristic of most images is that the principles, and classes of compression and various algorithm of image neighboring pixels are correlated and therefore contain compression. This paper attempts to give a recipe for selecting one of redundant information. The foremost task then is to find less the popular image compression algorithms based on Wavelet, correlated representation of the image. Two fundamental JPEG/DCT, VQ, and Fractal approaches. We review and discuss the components of compression are redundancy and irrelevancy advantages and disadvantages of data compression techniques and reduction. Redundancy reduction aims at removing algorithms for compressing grayscale images; give an experimental duplication from the signal source (image/video). Irrelevancy comparison on 256×256 commonly used image of Lenna and one reduction omits parts of the signal that will not be noticed by 400×400 fingerprint image. the signal receiver, namely the Human Visual System (HVS). In general, three types of redundancy can be identified: Keywords - Data compression, Image compression, JPEG, DCT, VQ, Wavelet, Fractal. A. Coding Redundancy I. INTRODUCTION A code is a system of symbols (letters, numbers, bits, and the like) used to represent a body of information or set of events. Data compression is one of many technologies that enables Each piece of information or events is assigned a sequence of today’s information revolution. Lossless data compression is code symbols, called a code word. The number of symbols in used to compact files or data into a smaller form. It is often each code word is its length. The 8-bit codes that are used to used to package up software before it is sent over the Internet or represent the intensities in the most 2-D intensity arrays downloaded from a web site to reduce the amount of time and contain more bits than are needed to represent the intensities. bandwidth required to transmit the data. Lossless data compression has the constraint that when data is uncompressed, B. Spatial Redundancy and Temporal Redundancy it must be identical to the original data that was compressed. Because the pixels of most 2-D intensity arrays are correlated Graphics, audio, and video compression such as JPEG, MP3, spatially, information is unnecessarily replicated in the and MPEG on the other hand use lossy compression schemes representations of the correlated pixels. In video sequence, which throw away some of the original data to compress the temporally correlated pixels also duplicate information. files even further. Image compression is the application of data compression on digital images. In effect, the objective is to C. Irrelevant Information reduce redundancy of the image data in order to be able to store Most 2-D intensity arrays contain information that is ignored or transmit data in an efficient form Uncompressed multimedia by the human visual system and extraneous to the intended (graphics, audio and video) data requires considerable storage use of the image. It is redundant in the sense that it is not capacity and transmission bandwidth. Despite rapid progress in used. Image compression research aims at reducing the mass-storage density, processor speeds, and digital number of bits needed to represent an image by removing the communication system performance, demand for data storage spatial and spectral redundancies as much as possible. capacity and data- transmission bandwidth continues to outstrip the capabilities of available technologies. The recent growth of data intensive multimedia-based web applications have not only [email protected] * Corresponding Author Email-Id sustained the need for more efficient ways to encode signals iJARS/ Vol. II/ Issue 2/Feb, 2013/329 1 http://www.ijars.in International Journal of Applied Research & Studies ISSN 2278 – 9480 Bits/ Uncompres Transmissio Multimedia Size/ Pixel or sed Size (B n Transmiss B. Predictive vs. Transform coding Data Duration Bits/ for bytes) Bandwidth ion Time In predictive coding, information already sent or available Sample (b for bits) is used to predict future values, and the difference is coded. Since this is done in the image or spatial domain, it is A page of 11"x8.5 Varyi 4.8KB 32.64 1.1 - 2.2 relatively simple to implement and is readily adapted to local text " ng Kb/page sec image characteristics. Differential Pulse Code Modulation resolut (DPCM) is one particular example of predictive coding. ion Transform coding, on the other hand, first transforms the Telephone 10 sec 8 bps 80KB 64 Kb/sec 22.2 sec image from its spatial domain representation to a different quality type of representation using some well-known transform and speech then codes the transformed values (coefficients). This method Grayscale 512x512 8 bpp 262KB 2.1 1 min provides greater data compression compared to predictive Image Mb/Image 13 sec methods, although at the expense of greater computation. Color 512x512 24 786KB 6.29 3 min Image bpp Mb/Image 39 sec V. CONVENTIONAL UNIVERSAL DATA COMPRESSION Medical 2048x16 12 5.16MB 41.3 23 min Image 80 bpp Mb/Image 54 sec SHD Image 2048x20 24 12.58 100 58 min Most universal compression techniques collect statistical 48 bpp MB Mb/Image 15 sec information to identify the highest frequency patterns in the data. In this section some of the previous attempts to address the universal compression problem by utilising statistical approaches are reviewed. III. NEED OF COMPRESSION The Table 1 show the qualitative transition from simple text A. Adaptive Compression Algorithm to full-motion video data and the disk space transmission Typically data compression algorithms have been bandwidth, and transmission time needed to store and implemented through a two-pass approach. In the first pass, transmit such uncompressed data. the algorithm collects information regarding the data to be compressed, such as the frequencies of characters or Table 1: substrings. In the second pass, the actual encoding takes place. Multimedia data types and uncompressed storage space, transmission A fixed encoding scheme is required to ensure that the bandwidth, and transmission time required. The prefix kilo- denotes a factor decoder is able to retrieve the original message. This approach of 1000 rather than 1024. has been improved through so called adaptive compression, where the algorithm only needs one pass to compress the data The examples given in the Table I clearly illustrate the need [25]. The main idea of adaptive compression algorithms is that for sufficient storage space, large transmission bandwidth, and the encoding scheme changes as the data being compressed. long transmission time for image, audio, and video data. Thus, the encoding of the nth symbol is based on the characteristics of the data until position n�1 [25]. A key At the present state of technology, the only solution is to advantage of adaptive compression is that it does not require compress multimedia data before its storage and transmission, the entire message to be loaded into the memory before the and decompress it at the receiver for play back. For example, compression process can start. with a compression ratio of 32:1, the space, bandwidth, and transmission time requirements can be reduced by a factor of Adaptive Huffman coding [26] is a statistical lossless 32, with acceptable quality. compression where the code is represented using a binary tree structure. This algorithm gives the most frequent characters IV. Different classes of compression techniques that appear in the file to be compressed shorter codes than those characters which are less frequent. Top nodes in the tree Two ways of classifying compression techniques are store the most frequent characters. To produce the compressed mentioned here. version of a file, the algorithm simply traverses the tree from A. Lossless vs. Lossy compression the root node to the target character. At each step the In lossless compression schemes, the reconstructed image, algorithm will encode 1 if it has moved to the left branch and after compression, is numerically identical to the original 0 otherwise. The Huffman tree creates a unique variable image. However lossless compression can only achieve a length code for each character. The novelty of the Adaptive modest amount of compression. An image reconstructed Huffman compression is that nodes swap locations within the following lossy compression contains degradation relative to tree during the compression. This allows the algorithm to the original. Often this is because the compression scheme adapt its compression as the frequency of the symbols changes completely discards redundant information. However, lossy throughout the file. schemes are capable of achieving much higher compression. Under normal viewing conditions, no visible loss is perceived Adaptive Arithmetic Coding (AC) is a statistical compression (visually lossless). iJARS/ Vol. II/ Issue 2/Feb, 2013/329 2 http://www.ijars.in International Journal of Applied Research & Studies ISSN 2278 – 9480 that learns the distribution of the source during the this model has similar performance to PPM it is not widely compression process [23]. AC has historic significance as it used. was the best alternative to Huffman coding after a gap of 25 years [23]. Adaptive AC encodes the source message to Lempel and Ziv developed a universal compression system in variable-length code in such a way that frequently used 1977 known as LZ77 [28].
Load more

Recommended publications
  • Word-Based Text Compression
    WORD-BASED TEXT COMPRESSION Jan Platoš, Jiří Dvorský Department of Computer Science VŠB – Technical University of Ostrava, Czech Republic {jan.platos.fei, jiri.dvorsky}@vsb.cz ABSTRACT Today there are many universal compression algorithms, but in most cases is for specific data better using specific algorithm - JPEG for images, MPEG for movies, etc. For textual documents there are special methods based on PPM algorithm or methods with non-character access, e.g. word-based compression. In the past, several papers describing variants of word- based compression using Huffman encoding or LZW method were published. The subject of this paper is the description of a word-based compression variant based on the LZ77 algorithm. The LZ77 algorithm and its modifications are described in this paper. Moreover, various ways of sliding window implementation and various possibilities of output encoding are described, as well. This paper also includes the implementation of an experimental application, testing of its efficiency and finding the best combination of all parts of the LZ77 coder. This is done to achieve the best compression ratio. In conclusion there is comparison of this implemented application with other word-based compression programs and with other commonly used compression programs. Key Words: LZ77, word-based compression, text compression 1. Introduction Data compression is used more and more the text compression. In the past, word- in these days, because larger amount of based compression methods based on data require to be transferred or backed-up Huffman encoding, LZW or BWT were and capacity of media or speed of network tested. This paper describes word-based lines increase slowly.
    [Show full text]
  • The Basic Principles of Data Compression
    The Basic Principles of Data Compression Author: Conrad Chung, 2BrightSparks Introduction Internet users who download or upload files from/to the web, or use email to send or receive attachments will most likely have encountered files in compressed format. In this topic we will cover how compression works, the advantages and disadvantages of compression, as well as types of compression. What is Compression? Compression is the process of encoding data more efficiently to achieve a reduction in file size. One type of compression available is referred to as lossless compression. This means the compressed file will be restored exactly to its original state with no loss of data during the decompression process. This is essential to data compression as the file would be corrupted and unusable should data be lost. Another compression category which will not be covered in this article is “lossy” compression often used in multimedia files for music and images and where data is discarded. Lossless compression algorithms use statistic modeling techniques to reduce repetitive information in a file. Some of the methods may include removal of spacing characters, representing a string of repeated characters with a single character or replacing recurring characters with smaller bit sequences. Advantages/Disadvantages of Compression Compression of files offer many advantages. When compressed, the quantity of bits used to store the information is reduced. Files that are smaller in size will result in shorter transmission times when they are transferred on the Internet. Compressed files also take up less storage space. File compression can zip up several small files into a single file for more convenient email transmission.
    [Show full text]
  • Dictionary Based Compression for Images
    INTERNATIONAL JOURNAL OF COMPUTERS Issue 3, Volume 6, 2012 Dictionary Based Compression for Images Bruno Carpentieri Ziv and Lempel in [1]. Abstract—Lempel-Ziv methods were original introduced to By limiting what could enter the dictionary, LZ2 assures compress one-dimensional data (text, object codes, etc.) but recently that there is at most one instance for each possible pattern in they have been successfully used in image compression. the dictionary. Constantinescu and Storer in [6] introduced a single-pass vector Initially the dictionary is empty. The coding pass consists of quantization algorithm that, with no training or previous knowledge of the digital data was able to achieve better compression results with searching the dictionary for the longest entry that is a prefix of respect to the JPEG standard and had also important computational a string starting at the current coding position. advantages. The index of the match is transmitted to the decoder using We review some of our recent work on LZ-based, single pass, log N bits, where N is the current size of the dictionary. adaptive algorithms for the compression of digital images, taking into 2 account the theoretical optimality of these approach, and we A new pattern is introduced into the dictionary by experimentally analyze the behavior of this algorithm with respect to concatenating the current match with the next character that the local dictionary size and with respect to the compression of bi- has to be encoded. level images. The dictionary of LZ2 continues to grow throughout the coding process. Keywords—Image compression, textual substitution methods.
    [Show full text]
  • A Survey on Different Compression Techniques Algorithm for Data Compression Ihardik Jani, Iijeegar Trivedi IC
    International Journal of Advanced Research in ISSN : 2347 - 8446 (Online) Computer Science & Technology (IJARCST 2014) Vol. 2, Issue 3 (July - Sept. 2014) ISSN : 2347 - 9817 (Print) A Survey on Different Compression Techniques Algorithm for Data Compression IHardik Jani, IIJeegar Trivedi IC. U. Shah University, India IIS. P. University, India Abstract Compression is useful because it helps us to reduce the resources usage, such as data storage space or transmission capacity. Data Compression is the technique of representing information in a compacted form. The actual aim of data compression is to be reduced redundancy in stored or communicated data, as well as increasing effectively data density. The data compression has important tool for the areas of file storage and distributed systems. To desirable Storage space on disks is expensively so a file which occupies less disk space is “cheapest” than an uncompressed files. The main purpose of data compression is asymptotically optimum data storage for all resources. The field data compression algorithm can be divided into different ways: lossless data compression and optimum lossy data compression as well as storage areas. Basically there are so many Compression methods available, which have a long list. In this paper, reviews of different basic lossless data and lossy compression algorithms are considered. On the basis of these techniques researcher have tried to purpose a bit reduction algorithm used for compression of data which is based on number theory system and file differential technique. The statistical coding techniques the algorithms such as Shannon-Fano Coding, Huffman coding, Adaptive Huffman coding, Run Length Encoding and Arithmetic coding are considered.
    [Show full text]
  • Enhanced Data Reduction, Segmentation, and Spatial
    ENHANCED DATA REDUCTION, SEGMENTATION, AND SPATIAL MULTIPLEXING METHODS FOR HYPERSPECTRAL IMAGING LEANNA N. ERGIN Bachelor of Forensic Chemistry Ohio University June 2006 Submitted in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY IN BIOANALYTICAL CHEMISTRY at the CLEVELAND STATE UNIVERSITY July 19th, 2017 We hereby approve this dissertation for Leanna N. Ergin Candidate for the Doctor of Philosophy in Clinical-Bioanalytical Chemistry degree for the Department of Chemistry and CLEVELAND STATE UNIVERSITY College of Graduate Studies ________________________________________________ Dissertation Committee Chairperson, Dr. John F. Turner II ________________________________ Department/Date ________________________________________________ Dissertation Committee Member, Dr. David W. Ball ________________________________ Department/Date ________________________________________________ Dissertation Committee Member, Dr. Petru S. Fodor ________________________________ Department/Date ________________________________________________ Dissertation Committee Member, Dr. Xue-Long Sun ________________________________ Department/Date ________________________________________________ Dissertation Committee Member, Dr. Yan Xu ________________________________ Department/Date ________________________________________________ Dissertation Committee Member, Dr. Aimin Zhou ________________________________ Department/Date Date of Defense: July 19th, 2017 Dedicated to my husband, Can Ergin. ACKNOWLEDGEMENT I would like to thank my advisor,
    [Show full text]
  • Answers to Exercises
    Answers to Exercises A bird does not sing because he has an answer, he sings because he has a song. —Chinese Proverb Intro.1: abstemious, abstentious, adventitious, annelidous, arsenious, arterious, face- tious, sacrilegious. Intro.2: When a software house has a popular product they tend to come up with new versions. A user can update an old version to a new one, and the update usually comes as a compressed file on a floppy disk. Over time the updates get bigger and, at a certain point, an update may not fit on a single floppy. This is why good compression is important in the case of software updates. The time it takes to compress and decompress the update is unimportant since these operations are typically done just once. Recently, software makers have taken to providing updates over the Internet, but even in such cases it is important to have small files because of the download times involved. 1.1: (1) ask a question, (2) absolutely necessary, (3) advance warning, (4) boiling hot, (5) climb up, (6) close scrutiny, (7) exactly the same, (8) free gift, (9) hot water heater, (10) my personal opinion, (11) newborn baby, (12) postponed until later, (13) unexpected surprise, (14) unsolved mysteries. 1.2: A reasonable way to use them is to code the five most-common strings in the text. Because irreversible text compression is a special-purpose method, the user may know what strings are common in any particular text to be compressed. The user may specify five such strings to the encoder, and they should also be written at the start of the output stream, for the decoder’s use.
    [Show full text]
  • 1 Introduction
    HELSINKI UNIVERSITY OF TECHNOLOGY Faculty of Electronics, Communications, and Automation Department of Communications and Networking Le Wang Evaluation of Compression for Energy-aware Communication in Wireless Networks Master’s Thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Technology. Espoo, May 11, 2009 Supervisor: Professor Jukka Manner Instructor: Sebastian Siikavirta 2 HELSINKI UNIVERSITY OF TECHNOLOGY ABSTRACT OF MASTER’S THESIS Author: Le Wang Title: Evaluation of Compression for Energy-aware Communication in Wireless Networks Number of pages: 75 p. Date: 11th May 2009 Faculty: Faculty of Electronics, Communications, and Automation Department: Department of Communications and Networks Code: S-38 Supervisor: Professor Jukka Manner Instructor: Sebastian Siikavirta Abstract In accordance with the development of ICT-based communication, energy efficient communication in wireless networks is being required for reducing energy consumption, cutting down greenhouse emissions and improving business competitiveness. Due to significant energy consumption of transmitting data over wireless networks, data compression techniques can be used to trade the overhead of compression/decompression for less communication energy. Careless and blind compression in wireless networks not only causes an expansion of file sizes, but also wastes energy. This study aims to investigate the usages of data compression to reduce the energy consumption in a hand-held device. By con- ducting experiments as the methodologies, the impacts of transmission on energy consumption are explored on wireless interfaces. Then, 9 lossless compression algo- rithms are examined on popular Internet traffic in the view of compression ratio, speed and consumed energy. Additionally, energy consumption of uplink, downlink and overall system is investigated to achieve a comprehensive understanding of compression in wireless networks.
    [Show full text]
  • Error-Resilient Coding Tools in MPEG-4
    I Error-Resilient Coding Tools In MPEG-4 t By CHENG Shu Ling A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF THE MASTER OF PHILOSOPHY DIVISION OF INFORMATION ENGnvfEERING THE CHINESE UNIVERSITY OF HONG KONG July 1997 • A^^s^ p(:^S31^^^ 4c—^w -v^ UNIVERSITY JgJ ' %^^y^^ s�,s^^ %^r::^v€X '^<"nii^Ji.^i:^'5^^ Acknowledgement I would like to thank my supervisor Prof. Victor K. W. Wei for his invaluable guidance and support throughout my M. Phil, study. He has given to me many fruitful ideas and inspiration on my research. I have been the teaching assistant of his channel coding and multimedia coding class for three terms, during which I have gained a lot of knowledge and experience. Working with Prof. Wei is a precious experience for me; I learnt many ideas on work and research from him. I would also like to thank my fellow labmates, C.K. Lai, C.W. Lam, S.W. Ng and C.W. Yuen. They have assisted me in many ways on my duties and given a lot of insightful discussions on my studies. Special thanks to Lai and Lam for their technical support on my research. Last but not least, I want to say thank you to my dear friends: Bird, William, Johnny, Mok, Abak, Samson, Siu-man and Nelson. Thanks to all of you for the support and patience with me over the years. ii Abstract Error resiliency is becoming increasingly important with the rapid growth of the mobile systems. The channels in mobile communications are subject to fading, shadowing and interference and thus have a high error rate.
    [Show full text]
  • Distortionless Source Coding 2
    ECE 515 Information Theory Distortionless Source Coding 2 1 Huffman Coding • The length of Huffman codewords has to be an integer number of symbols, while the self- information of the source symbols is almost always a non-integer. • Thus the theoretical minimum message compression cannot always be achieved. • For a binary source with p(x1) = 0.1 and p(x2) = 0.9 – H(X) = .469 so the optimal average codeword length is .469 bits – Symbol x1 should be encoded to l1 = -log2(0.1) = 3.32 bits – Symbol x2 should be encoded to l2 = -log2(0.9) = .152 bits 2 Improving Huffman Coding • One way to overcome the redundancy limitation is to encode blocks of several symbols. In this way the per-symbol inefficiency is spread over an entire block. – N = 1: ζ = 46.9% N = 2: ζ = 72.7% N = 3: ζ = 80.0% • However, using blocks is difficult to implement as there is a block for every possible combination of symbols, so the number of blocks (and thus codewords) increases exponentially with their length. – The probability of each block must be computed. 3 Peter Elias (1923 – 2001) 4 Arithmetic Coding • Arithmetic coding bypasses the idea of replacing a source symbol (or groups of symbols) with a specific codeword. • Instead, a sequence of symbols is encoded to an interval in [0,1). • Useful when dealing with sources with small alphabets, such as binary sources, and alphabets with highly skewed probabilities. 5 Arithmetic Coding Applications • JPEG, MPEG-1, MPEG-2 – Huffman and arithmetic coding • JPEG2000, MPEG-4 – Arithmetic coding only • ZIP – prediction by partial matching (PPMd) algorithm • H.263, H.264 6 Arithmetic Coding • Lexicographic ordering j−1 • Cumulative probabilities Puji= ∑p( ) i=1 u11111 xx x P u2 xx 11 x 2 P 2 uKKNN xxKK x K P • The interval Pj to Pj+1 defines uj 7 Arithmetic Coding • A sequence of source symbols is represented by an interval in [0,1).
    [Show full text]
  • 13. Compression and Decompression
    13. Compression and Decompression Algorithms for data compression usually proceed as follows. They encode a text over some nite alphabet into a sequence of bits, hereby exploiting the fact that the letters of this alphabet occur with different frequencies. For instance, an e occurs more frequently than a q and will therefore be assigned a shorter codeword. The quality of the compression procedure is then measured in terms of the average codeword length. So the underlying model is probabilistic, namely we consider a nite alphabet and a probability distribution on this alphabet, where the probability distribution reects the (re- lative) frequencies of the letters. Such a pair an alphabet with a probability distribution is called a source. We shall rst introduce some basic facts from Information Theory. Most important is the notion of entropy, since the source entropy characterizes the achievable lower bounds for compressibility. The source model to be best understood, is the discrete memoryless source. Here the letters occur independently of each other in the text. The use of prex codes, in which no codeword is the beginning of another one, allows to compress the text down to the entropy of the source. We shall study this in detail. The lower bound is obtained via Kraft's inequality, the achievability is demonstrated by the use of Huffman codes, which can be shown to be optimal. There are some assumptions on the discrete memoryless source, which are not fullled in most practical situations. Firstly, usually this source model is not realistic, since the letters do not occur independently in the text.
    [Show full text]
  • Introduction to Data Compression, by Guy E. Blelloch
    Introduction to Data Compression∗ Guy E. Blelloch Computer Science Department Carnegie Mellon University blellochcs.cmu.edu January 31, 2013 Contents 1 Introduction 3 2 Information Theory 5 2.1 Entropy ........................................ 5 2.2 The Entropy of the English Language . ...... 6 2.3 Conditional Entropy and Markov Chains . ...... 7 3 Probability Coding 10 3.1 PrefixCodes...................................... 10 3.1.1 Relationship to Entropy . 11 3.2 HuffmanCodes .................................... 13 3.2.1 CombiningMessages............................. 15 3.2.2 Minimum Variance Huffman Codes . 15 3.3 ArithmeticCoding ................................. 16 3.3.1 Integer Implementation . 19 4 Applications of Probability Coding 22 4.1 Run-lengthCoding .................................. 25 4.2 Move-To-FrontCoding .............................. 26 4.3 ResidualCoding:JPEG-LS. 27 4.4 ContextCoding:JBIG ................................ 28 4.5 ContextCoding:PPM................................. 29 ∗This is an early draft of a chapter of a book I’m starting to write on “algorithms in the real world”. There are surely many mistakes, and please feel free to point them out. In general the Lossless compression part is more polished than the lossy compression part. Some of the text and figures in the Lossy Compression sections are from scribe notes taken by Ben Liblit at UC Berkeley. Thanks for many comments from students that helped improve the presentation. c 2000, 2001 Guy Blelloch 1 5 The Lempel-Ziv Algorithms 32 5.1 Lempel-Ziv 77 (Sliding Windows) . ...... 32 5.2 Lempel-Ziv-Welch ................................ 34 6 Other Lossless Compression 37 6.1 BurrowsWheeler ................................... 37 7 Lossy Compression Techniques 40 7.1 ScalarQuantization .............................. 41 7.2 VectorQuantization.............................. 41 7.3 TransformCoding.................................. 43 8 A Case Study: JPEG and MPEG 44 8.1 JPEG ........................................
    [Show full text]
  • Energy-Aware Lossless Data Compression
    Energy-Aware Lossless Data Compression KENNETH C. BARR and KRSTE ASANOVIC´ MIT Computer Science and Artificial Intelligence Laboratory Wireless transmission of a single bit can require over 1000 times more energy than a single 32-bit computation. It can therefore be beneficial to perform additional computation to reduce the number of bits transmitted. If the energy required to compress data is less than the energy required to send it, there is a net energy savings and an increase in battery life for portable computers. This article presents a study of the energy savings possible by losslessly compressing data prior to transmission. A variety of algorithms were measured on a StrongARM SA-110 processor. This work demonstrates that, with several typical compression algorithms, there is a actually a net energy increase when compression is applied before transmission. Reasons for this increase are explained and suggestions are made to avoid it. One such energy-aware suggestion is asymmetric compression, the use of one compression algorithm on the transmit side and a different algorithm for the receive path. By choosing the lowest-energy compressor and decompressor on the test platform, overall energy to send and receive data can be reduced by 11% compared with a well-chosen symmetric pair, or up to 57% over the default symmetric zlib scheme. Categories and Subject Descriptors: C.4 [Performance of Systems]—Performance attributes General Terms: Algorithms, Measurement, Performance Additional Key Words and Phrases: Compression, energy-aware, lossless, low-power, power-aware 1. INTRODUCTION Wireless communication is an essential component of mobile computing, but the energy required for transmission of a single bit has been measured to be over 1000 times greater than for a single 32-bit computation.
    [Show full text]