Silkroad Message Board - Msg: 10864024

Sponsor

To: Jay Lowe who wrote (468)	8/7/1999 4:15:00 PM
From: ftth	Read Replies (1) of 626

Hi Jay, long time no (virtual) see! You're quite right, lossy compression of any kind isn't appropriate for the network data streams, since they don't gracefully degrade with loss of data like an image or audio can (to a point of course). Lossy compression could only be applied to such "gracefully degrading" sources, early in their journey down the wire before any modulation or channel coding (if applicable). Aware's use of wavelets for compression is new to me--I'm not 'aware'<gg> of their work in this area. I was aware of their interest in modulation using the Discrete Wavelet Transform in a multicarrier system. Sort of a wavelet version of DMT (discrete multitone, which uses the Discrete Fourier Transform). They had proposed this for the IEEE802.14 advanced PHY (for cable data systems, upstream). Amati had proposed a true-to-form DMT system for same Advanced PHY. Zenith, BTW, had proposed an S-CDMA system for cable upstream in 802.14 PHY.