the concept of siblings is simple .Imagine a structure where u have the employee database .The employee datbase contains the fields like empname,empid and empdepno and field like that.
In IMS each of these fields is stored as the segment.
Let the employee details be the root segment i.e the parent segement.This parent segment has the sub segemnts called empname, empid and emdeptno.These segments have the same parent ie empdetails hence the segments empname, empid and empdeptno are called siblings.
To add more comments to Vijay Maduri,
siblings are also called as Segment occurrences.
Segment is the format how the data gets stored and segment occurrences / sibilings is the actual data stored in the format of the segment. and each sibling is linked by a pointer. These pointers can be Hirarchial forward pointer or back pointer or both. Such Information will be given, during DBD definition.