Currently i have column called TEXT in my table having data with "High SMS usage - VPMN: XXXXX - SRC: XXX - SDR: XX.YY - Will Report"
By using regular expression with delimiter(-) i can able to separate string like attached screen shot 
But i want in this way 
Below is the query i am using.
Please suggest me how can i use regular expressions to get my desired output.
Maybe you can use a combination of REGEXP_SUBSTR() and REPLACE().
Test data:
create table messages ( message_ )
as
select 'High SMS usage - VPMN: PHLGT - SRC: NRT - SDR: 22.64 - Will Report' from dual union all
select 'High SMS usage - VPMN: ABCDE - SRC: ABCD - SDR: 222.64 - Will Report' from dual union all
select 'High SMS usage - VPMN: FGHI - SRC: EFG - SDR: 2222.64 - Will Report' from dual union all
select 'High SMS usage - VPMN: JKL - SRC: HIJK - SDR: 222.64 - Will Report' from dual union all
select 'High SMS usage - VPMN: MNOPQR - SRC: LMN - SDR: 22.64 - Will Report' from dual ;
Query:
select
replace( regexp_substr( noblanks, '[^-]+', 1, 2), 'VPMN:' ) as VPMN
, replace( regexp_substr( noblanks, '[^-]+', 1, 3), 'SRC:' ) as DATA_SRC
, replace( regexp_substr( noblanks, '[^-]+', 1, 4), 'SDR:' ) as SDR_VALUE
, text_
from (
select
replace( message_, ' ' ) as noblanks -- message without blanks
, message_ as text_ -- original message
from messages
) ;
-- result
VPMN DATA_SRC SDR_VALUE TEXT_
PHLGT NRT 22.64 High SMS usage - VPMN: PHLGT - SRC: NRT - SDR: 22.64 - Will Report
ABCDE ABCD 222.64 High SMS usage - VPMN: ABCDE - SRC: ABCD - SDR: 222.64 - Will Report
FGHI EFG 2222.64 High SMS usage - VPMN: FGHI - SRC: EFG - SDR: 2222.64 - Will Report
JKL HIJK 222.64 High SMS usage - VPMN: JKL - SRC: HIJK - SDR: 222.64 - Will Report
MNOPQR LMN 22.64 High SMS usage - VPMN: MNOPQR - SRC: LMN - SDR: 22.64 - Will Report
Explanation: the inline view (subquery) gives us the message without blanks/spaces, and the original message text. We then use regexp_substr() to find the key-value pairs, and replace() to remove the "labels" (VPMN:, SRC:, SDR:). Dbfiddle here.
