Data Compression Algorithms for Energy-Constrained Devices in Delay Tolerant Networks Christopher Sadler Margaret Martonosi Princeton University

2 Why do we care about Compression in Sensor Networks? Short Range 125 m Med. Range 300 m Long Range 15 km Success = Energy Savings Energy = Compute Energy + Transmit Energy ~2 Million! ~32,000 ~4,000

3 Why do we care about Compression in Sensor Networks? Unreliability: Retransmission Extra energy cost Easier to amortize original energy cost of compression Medium Range Radio (CC1000) ~128,000! ~32,000

4 Why do we care about Compression in Sensor Networks? Source Local Energy Tradeoff:  Transmit all data vs.  Compress data  Store data  Transmit compressed data Sink Downstream Energy Tradeoff:  Relay all data vs.  Relay compressed data Downstream Energy Tradeoff:  Relay all data vs.  Relay compressed data Savings Accumulate with Hop Count

5 This work Lossless compression algorithms tailored to static and mobile sensors Generally applicable for WSNs  Aggregation  Spatial-temporal correlation Implementation and evaluation on the real data set  Great Duck Island Monitoring, ZebraNet, …

6 Outline Design Criteria and LZW Compression  What we want in Sensor Compression?  How do we adapt LZW to Sensors? Using Compression to Conserve Energy Conclusions

7 Sensor Network Compression: Energy Savings for Everyone Need a general purpose, lossless compression algorithm that can work across the design space ZebraNet (Princeton): Outdoors, Mobile Great Duck Island (UCB): Outdoors, Stationary SensorScope (EPFL): Indoors, Stationary

8 Sensor Network Compression: What do we want? Low Transmission Overhead Computationally Simple  Compute energy of compression should not outweigh transmission energy savings Bounded Memory Footprint  To fit in small sensor node memories Adaptive to general data sets  Exploit repetition in general input data streams  Work on small blocks of data

9 LZW Compression LZW is a dictionary-based algorithm which encodes new strings based on previously encountered strings. AAAABAAABCC (8bits * 11 = 88bits required) Standard Dic A = 65 B = 66 C = Z = ? = 255 Encoded String Output Stream New Dic entry A 65 AA = 256 AA AAA = 257 A AB = 258 B BA = 259 AAA AAAB = 260 B BC = 261 C CC = 262 C bits required

10 Outline Design Criteria and LZW Compression  What we want in Sensor Compression?  How do we adapt LZW to Sensors? Using Compression to Conserve Energy Conclusions

11 S-LZW: LZW for Sensor Nodes Dictionary decisions  How large should we make the dictionary?  What do we do if the dictionary fills? Data decisions  How much data should we compress at once?  Can we shape the dictionary to improve compression?  Can we shape the data to make it easier to compress?

12 S-LZW: LZW for Sensor Nodes Dictionary decisions  How large should we make the dictionary?  What do we do if the dictionary fills? Data decisions  How much data should we compress at once? Longer data streams => better compression learning But too long => high retransmit cost when packets dropped  Can we shape the dictionary to improve compression?  Can we shape the data to make it easier to compress? Examined by experiments

13 SENSOR DATA – N BYTES GENERATED OVER TIME … 528 B Block (2 Flash Pages) COMP. ALGORITHM Compressed Data Independent groups of 10 or fewer dependent packets …… … … … … 528 B Block (2 Flash Pages) S-LZW Idea 1: Data Size

14 S-LZW Idea 2: Mini-Caching Dictionary Tree Mini- Cache (N entries) 01 Escape Bit 9 Bit Entries10 Bit Entries(Log 2 N)+1 Bit Entries Exploit fine-grained locality even in short sensor data sequences Proposal: Mini-cache to tightly encode MRU entries  Hit => Saves multiple bits. Miss => Costs just 1 extra escape bit

15 S-LZW Idea 3: Data Transforms BWT – Reversible method of transforming data used in bzip2 and applicable for all data Proposal: Structured Transform - Create a matrix of readings and transpose it to create runs  Simple, but effective cb 4e … d5 4e … d8 4e … db 4e 2b 62 … cb d5 d8 db … 4e 4e 4e 4e … b … …......

16 Outline Design Criteria and LZW Compression Using Compression to Conserve Energy  Local and Downstream Energy  The Influence of Unreliable Communications  The Effects of Shaping the Data Conclusions

17 Measurement Methodology: CPU and Radios Evaluation Platform:  TI microcontroller MSP430: 10 kB RAM, 48 kB ROM  Off-Chip Flash: 4 Mbit Atmel  3 Radios: Short (CC2420), Medium (CC1000), and Long (XTend) range 3 Real World Datasets…  SensorScope (SS) – Indoor, Stationary  Great Duck Island (GDI) – Outdoor, Stationary  ZebraNet (ZNet) – Outdoor, Mobile … And one Compression Benchmark  Geo from the Calgary Corpus (Calgeo)

18 Local Energy Savings CC2420 (Short Range) XTend (Long Range) Data Compressed with S-LZW with Mini-Cache Model assumes 100% reliability 2.6X Gain 1.2X Gain 1.7X Gain 1% Loss 15+% Loss Lower is better

19 Downstream Energy Savings CC2420 (Short Range) XTend (Long Range) ZNet data, Compressed with S-LZW with Mini-Cache Model assumes 100% reliability

20 Coping with Unreliability 1.Energy savings increase linearly with hop count GDI Data, Compressed with S-LZW with Mini-Cache CC2420 (Short Range)

21 Coping with Unreliability 1.Energy savings increase linearly with hop count GDI Data, Compressed with S-LZW with Mini-Cache CC2420 (Short Range) 2.At a 90% success rate, we save energy locally

22 Transforms to Improve Performance BWT – Reversible method of transforming data through sorting it XTend (Long Range) We sort with an iterative quicksort algorithm, but this comes at a high computational cost Normalized against sending the data without compression 3.4X Gain 2.4X Gain 7-8% Improvement Top – Radio Energy Middle – Flash Energy Bottom – CPU Energy Model assumes 100% reliability

23 Radio Range? Number of Hops? Data Composition? Number of Hops? StructuredGeneral Radio Range? S-LZW-MC8-ST ~2.3X Savings RLE-ST ~1.9X Savings S-LZW-MC8-ST ~2.4X Savings FewMany Short Medium/ Long S-LZW-MC16-BWT ~1.8X Savings S-LZW-MC32 ~1.4X Savings S-LZW-MC16-BWT ~1.8X Savings FewMany Short Medium/ Long Algorithms Summary MC – Mini-Cache ST – Structured Transform BWT – Unstructured Transform

24 Conclusions Existing techniques (LZW, Cache, BWT …) applied to new application domains (WSNs)  Less novelty and originality  Incremental work  Upper bound 4 points Systematic solution with complete considerations  Dictionary size, data size, local cache, data transformation … Substantial merits proved by sound experiments  Different types of radio transceivers (CC2420, CC1000, XTend)  Real-world data set