Ideas for Corebaseit

My recommended publishing order

I would structure the next wave like this:

  1. Nyquist Is Not Shannon - ready to review | ✅ | posted to corebaseit | ✅ | posted to linkedln | |
  2. SNR: The Number That Decides Whether a Signal Survives - ready to review | ✅ | posted to corebaseit | | posted to linkedln | |
  3. Why Your Internet Has a Speed Limit - ready to review | ✅ | posted to corebaseit | | posted to linkedln | |
  4. QPSK Explained - ready to review | | posted to corebaseit | | posted to linkedln | |
  5. Matched Filtering - ready to review | | posted to corebaseit | | posted to linkedln | |
  6. Equalizers - ready to review | | posted to corebaseit | | posted to linkedln | |
  7. OFDM Explained - ready to review | | posted to corebaseit | | posted to linkedln | |
  8. Source Coding vs. Channel Coding - ready to review | | posted to corebaseit | | posted to linkedln | |
  9. Error Correction - ready to review | | posted to corebaseit | | posted to linkedln | |
  10. Stochastic Is Not Magic - ready to review | | posted to corebaseit | | posted to linkedln | |

That gives you a clean learning path:

sampling → noise → capacity → modulation → receiver design → channel correction → modern systems → information theory → AI connection

The strongest next post, in my opinion, is:

Nyquist Is Not Shannon: Two Limits Every DSP Engineer Should Keep Separate

It builds directly on your current content, avoids repetition, and gives readers a clear “aha” moment.

Based on what you already have, your current thread is very strong around:

  • Wiener/MMSE
  • LMS/adaptive filtering
  • SGD connection to AI
  • FIR design / FPGA
  • entropy, stochastic systems, and information theory

Your category pages currently show 3 Digital Communications posts and 4 Information Theory posts, with overlap around adaptive filters and MMSE. That gives you a good foundation for a series that moves from signals → noise → filters → learning → communication limits.  

Here are post ideas that would add value and naturally build on what you already published.

  1. Why Your Internet Has a Speed Limit: Shannon Capacity Explained Without the Hype

This would be a perfect next pillar post.

You already introduced entropy and stochastic systems. The natural next step is channel capacity: why bandwidth, noise, and signal power define the theoretical ceiling of communication.

Possible angle:

We often blame routers, software, or providers for slow internet. But underneath every communication system sits a deeper limit: how much information a noisy physical channel can carry.

Core concepts:

  • Shannon capacity
  • Bandwidth
  • Signal-to-noise ratio
  • Why “more power” is not always the answer
  • Why fiber, Wi-Fi, 5G, and DSL are all capacity-limited systems

Suggested title:

Why Your Internet Has a Speed Limit: Shannon Capacity and the Physics of Information

  1. SNR: The Quiet Number Behind Every Communication System

This would bridge nicely between digital communications and information theory.

SNR is one of those concepts that appears everywhere: RF, Wi-Fi, DSL, optical links, modems, ADCs, ML signal pipelines, even payments terminals with NFC/contactless constraints.

Possible angle:

Before a receiver can decode anything, it must first answer a brutal question: is the signal stronger than the randomness around it?

Core concepts:

  • Signal power vs. noise power
  • Decibels
  • Thermal noise
  • Why higher data rates require better SNR
  • Why poor SNR causes retries, lower modulation order, or link failure

Suggested title:

SNR: The Number That Decides Whether a Signal Survives

  1. From Nyquist to Shannon: Two Limits Engineers Should Not Confuse

This would be very valuable because people often mix these concepts.

You have already discussed filtering, ADCs, aliasing, and DSP diagrams recently. This post would connect beautifully with that.

Possible angle:

Nyquist tells us how fast we must sample to avoid losing the shape of a signal. Shannon tells us how much information a noisy channel can carry. They are related, but they are not the same limit.

Core concepts:

  • Nyquist sampling theorem
  • Aliasing
  • Anti-aliasing filters
  • Shannon capacity
  • Sampling limit vs. information limit

Suggested title:

Nyquist Is Not Shannon: Two Limits Every DSP Engineer Should Keep Separate

  1. Why Filters Are Not Just “Cleaning Tools”

This would build directly on your FIR, Wiener, and adaptive filter posts.

Possible angle:

We often describe filters as tools that remove noise. But in real systems, filters shape what the receiver is allowed to believe.

Core concepts:

  • FIR vs. IIR
  • Matched filtering
  • Wiener filtering
  • Anti-aliasing
  • Equalization
  • Filtering as information selection, not just noise removal

Suggested title:

Filters Do More Than Remove Noise: They Decide What Information Survives

  1. Matched Filters: The Receiver’s Best Guess Before the Decision

This is an excellent follow-up after MMSE/Wiener and before modulation/demodulation.

Possible angle:

A matched filter does not magically know the transmitted bit. It simply maximizes the chance that the right structure rises above the noise at the decision instant.

Core concepts:

  • Pulse shaping
  • Symbol timing
  • Integrate-and-dump
  • Maximizing SNR
  • Receiver decision point
  • Why matched filtering appears in modems, radar, and digital communication

Suggested title:

Matched Filtering: How Receivers Find Meaning Inside Noise

  1. QPSK Explained: How Two Bits Become One Symbol

You have discussed QPSK and symbols recently. This would be a very accessible post for engineers crossing from software into communications.

Possible angle:

In software, two bits are just two bits. In a modem, those two bits become a point in a plane, a phase change, and eventually a decision under uncertainty.

Core concepts:

  • Bits vs. symbols
  • I/Q plane
  • Constellation diagrams
  • Noise movement around constellation points
  • Decision boundaries
  • Why higher modulation orders need better SNR

Suggested title:

QPSK Explained: When Bits Become Geometry

  1. Equalizers: Teaching a Receiver to Undo the Channel

This fits perfectly after LMS/adaptive filters.

Possible angle:

A communication channel does not simply carry a signal. It distorts it. Equalization is the receiver’s attempt to learn the inverse personality of the channel.

Core concepts:

  • Inter-symbol interference
  • Multipath
  • Channel impulse response
  • LMS equalizer
  • Training sequences
  • Adaptive correction

Suggested title:

Equalizers: How Receivers Learn to Undo Distortion

  1. Echo Cancellers: The Adaptive Filter You Use Every Day

This is very strong because it connects telecom DSP with something readers understand immediately.

Possible angle:

Echo cancellation is one of the cleanest examples of adaptive filtering in production: a system that listens, predicts its own echo, subtracts it, and keeps learning.

Core concepts:

  • Acoustic echo
  • Line echo
  • LMS/NLMS
  • Reference signal
  • Error signal
  • Real-time adaptation

Suggested title:

Echo Cancellers: Adaptive Filters Hiding in Plain Sight

  1. OFDM: Why Modern Communication Splits One Fast Stream Into Many Slow Ones

This would be a high-value digital communications post. It connects to Wi-Fi, DSL, 4G, 5G, and optical/wireline systems.

Possible angle:

OFDM looks strange at first: instead of sending one fast stream, we split the data across many slower subcarriers. The trick is that this makes the channel easier to survive.

Core concepts:

  • Subcarriers
  • Orthogonality
  • FFT/IFFT
  • Multipath resilience
  • Cyclic prefix
  • Why OFDM dominates modern communication

Suggested title:

OFDM Explained: Why Modern Networks Break Data Into Many Tiny Channels

  1. Error Correction: Why Reliable Communication Sends More Than the Message

This would extend your information theory category and eventually connect to payments reliability/security ideas.

Possible angle:

A good communication system does not simply transmit data. It transmits enough structure for the receiver to notice and repair damage.

Core concepts:

  • Redundancy
  • Parity
  • Hamming distance
  • Convolutional codes
  • Reed-Solomon
  • LDPC
  • Turbo codes
  • Why redundancy is not waste

Suggested title:

Error Correction: Why Reliable Systems Add Redundancy on Purpose

  1. Source Coding vs. Channel Coding: Compression and Protection Are Opposite Forces

This one would be excellent for your Information Theory category.

Possible angle:

Compression removes redundancy. Channel coding adds redundancy. At first, that sounds contradictory. In reality, they solve different problems.

Core concepts:

  • Entropy
  • Compression
  • Removing predictable structure
  • Adding controlled redundancy
  • Source coding theorem
  • Channel coding theorem

Suggested title:

Compression Removes Redundancy. Error Correction Adds It. Both Are Right.

  1. From Thermal Noise to AI Randomness: Why “Stochastic” Does Not Mean “Magic”

This would connect your AI and information theory themes.

Possible angle:

Engineers used stochastic models long before modern AI. Randomness was never magic; it was a practical way to reason about systems too complex to predict exactly.

Core concepts:

  • Thermal noise
  • Stochastic processes
  • Probability distributions
  • SGD
  • Sampling
  • Diffusion models
  • LLM generation
  • Engineering vs. AI vocabulary

Suggested title:

Stochastic Is Not Magic: From Thermal Noise to Modern AI