A distributed port scan operates by:

Blocking access to scanning clients describes a defensive countermeasure taken by the target, not a characteristic of how distributed scanning is performed.

Using denial-of-service software against TCP ports describes a DoS attack, which is a distinct offensive category unrelated to port reconnaissance.

Scanning clients probing a target have no mechanism or reason to block access to that same target, making this option technically contradictory.

In a distributed port scan, multiple machines each perform a portion of the total scan - for example one host scans ports 1-1024 while another scans 1025-2048 - so no single source generates enough traffic to trigger rate-based IDS or firewall rules. The results from all participating scanners are then aggregated and correlated to produce a complete picture of the target. This technique defeats threshold-based detection by keeping each individual scanner's activity below alert levels.