Lossless Compression of Cloud-Cover Forecasts for Low-Overhead Distribution in Solar-Harvesting Sensor Networks Christian Renner and Phu Anh Tuan Nguyen ENSsys 14, Memphis, TN, USA November 6 th, 2014
Motivation Making the Most of Harvestable Energy Objective perpetual operation Challenge changing weather conditions Instrument load adaptation Prerequisite energy harvest prediction 1
Motivation Sources of Harvest Variation Global Weather conditions Inter and intra-day variation Seasonal effects Distribution required Unused Local Location-specific pattern Shades of buildings and trees Dirt deposits Hardware aging Locally available Frequently used 2
Motivation Advantages of Integrating Cloud Cover Forecasts harvest (ma) ME (ma) 12 8 4 0 2 1 0 1 EWMA Kimball-1 2 70 71 72 73 74 75 76 77 78 79 80 3
Motivation Data Distribution Challenge One-to-many communication (data distribution) Reversed data flow data collection Elevated network energy expenditure Approach Piggy-back on collection data acknowledgments Reduces energy overhead Requires small size compression 4
1 2 3 4 Lossless Compression
Lossless Compression Compression Checklist Complete forecasts differential data problematic in case of packet loss Lossless compression coarse-grained input data, no experience with lossy forecasts Light-weight decoding execution on resource-constrained sensor nodes, decoding consumes energy Exploit data properties and patterns 5
Lossless Compression Cloud-Cover Forecasts cloud cover (eighth) 8 6 4 2 0 0 6 12 18 24 30 36 42 48 54 60 66 72 time (h) Cloud cover in eighths Hourly resolution Multi-day horizon Hourly updates 6
Lossless Compression Data Analysis 30 100 frequency (%) 20 10 frequency (%) 80 60 40 20 0 0 1 2 3 4 5 6 7 cloud cover (eighth) 0 3 2 1 0 1 2 3 cloud cover delta (eighth) difference of adjacent forecast values (deltas) is zero in 82% of all cases at most one in 99% of all cases deltas have lower entropy than absolute values 7
Lossless Compression Compression Algorithms Delta Coding (DC) Huffman encoding of delta values (-8,..., 8) First value unencoded (absolute value) Cloud-cover forecast 5 5 6 5 5 5 8 8 Deltas 5 0 +1-1 0 0 +3 0 Encoding 0101 0 10 110 0 0 1111110 0 8
Lossless Compression Compression Algorithms Partial Delta Coding (PDC) Huffman encoding of delta values -1, 0, 1 Larger deltas encoded as absolute values with prefix First value unencoded (absolute value) Cloud-cover forecast 5 5 6 5 5 5 8 8 Deltas 5 0 +1-1 0 0 +3 0 Encoding 0101 0 10 110 0 0 111 1000 0 9
Lossless Compression Compression Algorithms Daylight Delta Coding (DDC) PDC encoding Omit night slots Indicate number of values before sunrise / sunset Cloud-cover forecast 5 5 6 5 5 5 8 8 Deltas 5 0 0 +3 0 Encoding 0101 0 0 111 1000 0 10
1 2 3 4 Evaluation
Evaluation Setup and Metrics Data set Cloud-cover forecasts from an online weather forecast service April 2013 to July 2014 10 590 forecasts Hourly resolution, 10-day horizon Sunrise and sunset times from sunrise equation Compression DC, PDC, DDC Message sizes (including all static information) 11
Evaluation Results Compression Performance and Comparison 40 DC 40 DDC encoded length (B) 30 20 10 encoded length (B) 30 20 10 0 1 2 3 4 5 6 7 forecast horizon (d) 0 1 2 3 4 5 6 7 forecast horizon (d) 12
Evaluation Results Optimality Study horizon (d) avg. compression code length (bit/value) optimal DC PDC DDC 1 1.00 1.50 1.51 1.43 2 0.94 1.41 1.42 1.14 3 0.92 1.37 1.38 1.04 4 0.91 1.35 1.36 0.99 5 0.90 1.34 1.35 0.95 6 0.90 1.32 1.33 0.93 7 0.89 1.31 1.32 0.91 13
Evaluation Results Seasonal Daytime Influence 3-day forecasts: saving of DDC over PDC 50 saving (%) 0 50 Apr May Jun Jul Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun Jul time of year 14
Evaluation Summary Results DDC achieves best results Compression of up to 76% Forecasts of up to 3 d consume at most 17 byte Efficient (decoding) implementation 15
1 2 3 4 Conclusion & Future Work
Conclusion & Future Work Conclusion Cloud-cover forecasts improve (solar) harvest prediction accuracy are compressible to a few bytes with low computation overhead can be distributed into the network via acknowledgment piggy-backing 16
Conclusion & Future Work Current and Next Steps Implementation for TinyOS Latency evaluation for piggy-backed data distribution Field test Compression / distribution of other weather metrics e.g., sunshine duration per hour 17
Christian Renner Research Assistant and Lecturer Fon +49 / (0)451 500 3741 E-Mail renner@iti.uni-luebeck.de http://www.iti.uni-luebeck.de Lossless Compression of Cloud-Cover Forecasts for Low-Overhead Distribution in Solar-Harvesting Sensor Networks Christian Renner and Phu Anh Tuan Nguyen ENSsys 14, Memphis, TN, USA November 6 th, 2014