50,000 Page Views in 2020

And counting …

Last evening the CSP Blog crossed the 50,000 page-view threshold for 2020, a yearly total that has not been achieved previously!

I want to thank each reader, each commenter, and each person that’s clicked the Donate button. You’ve made the CSP Blog the success it is, and I am so grateful for the time you spend here.

On these occasions I put some of the more interesting CSP-Blog statistics below the fold. If you have been wanting to see a post on a particular CSP or Signal Processing ToolKit topic, and it just hasn’t appeared, feel free to leave me a note in the Comments section.

2020 continued the trend of year-over-year increase in page views:

2019’s total was just shy of 50,000 at 49,917.

November 2020 broke the 5000-views barrier for the first time:

Earlier this year, my most-popular post (no, it isn’t about machine learning, thankfully, it is about The Spectral Correlation Function) broke the 10,000-views barrier, and the second-most-popular post is not far behind:

Author: Chad Spooner

I'm a signal processing researcher specializing in cyclostationary signal processing (CSP) for communication signals. I hope to use this blog to help others with their cyclo-projects and to learn more about how CSP is being used and extended worldwide.

2 thoughts on “50,000 Page Views in 2020”

  1. I’m looking forward to your next modulation recognizing(MR) poll.

    I read your publication [25],[26], and [28]. I want to see more details on the calibration process that estimate the power |A|^2. Is this process done by regression with scaling parameter A?

    Based on my understanding, the cyclostationarity-base MR method relies on the cyclic-cumulant. Is the cyclic-cumulant computed at the specific time delay for example [0,0,…,0] fo squared-root raised cosine?

    BTW, I tried high-order cyclic cumulant on GPU, really computational heavy even for two variable [t1,t2,0,0,…,0] due to heavy recursion for order greater than 6.

    1. For estimating A, or A^2, in the context of cyclic-cumulant MR, consider a least-squares estimate using the estimated set of cyclic cumulants and an ideal set corresponding to a signal with unit power.

      The delay-reduced classifier I talk about in [26] typically can work with just a single delay vector, the origin. But it will depend on the signals of interest in the classification problem.

Leave a Comment, Ask a Question, or Point out an Error