USE tempdb;
GO

-- creo una tabella simile alla tua per prova
CREATE TABLE dbo.position
(
    position_id int NOT NULL
  , location_id int NULL
  , PLN_ID int NOT NULL
  , fixel_id int NOT NULL
  , CONSTRAINT PK_dboposition PRIMARY KEY CLUSTERED
  (
      position_id
    , PLN_ID
    , fixel_id
  )  
)
GO

-- valori di esempio (c'è anche il 202)
INSERT INTO dbo.position (position_id, location_id, PLN_ID, fixel_id)
VALUES (1, NULL, 10, 10)
INSERT INTO dbo.position (position_id, location_id, PLN_ID, fixel_id)
VALUES (1, NULL, 20, 20)
INSERT INTO dbo.position (position_id, location_id, PLN_ID, fixel_id)
VALUES (1, NULL, 30, 30)
INSERT INTO dbo.position (position_id, location_id, PLN_ID, fixel_id)
VALUES (2, NULL, 10, 10)
INSERT INTO dbo.position (position_id, location_id, PLN_ID, fixel_id)
VALUES (2, NULL, 20, 20)
INSERT INTO dbo.position (position_id, location_id, PLN_ID, fixel_id)
VALUES (202, NULL, 10, 10)
INSERT INTO dbo.position (position_id, location_id, PLN_ID, fixel_id)
VALUES (202, NULL, 20, 20)
INSERT INTO dbo.position (position_id, location_id, PLN_ID, fixel_id)
VALUES (202, NULL, 30, 30)
INSERT INTO dbo.position (position_id, location_id, PLN_ID, fixel_id)
VALUES (202, NULL, 40, 40)
INSERT INTO dbo.position (position_id, location_id, PLN_ID, fixel_id)
VALUES (202, NULL, 50, 50)
INSERT INTO dbo.position (position_id, location_id, PLN_ID, fixel_id)
VALUES (4, NULL, 10, 10)
INSERT INTO dbo.position (position_id, location_id, PLN_ID, fixel_id)
VALUES (4, NULL, 20, 20)
INSERT INTO dbo.position (position_id, location_id, PLN_ID, fixel_id)
VALUES (4, NULL, 30, 30)
INSERT INTO dbo.position (position_id, location_id, PLN_ID, fixel_id)
VALUES (5, NULL, 10, 10)
INSERT INTO dbo.position (position_id, location_id, PLN_ID, fixel_id)
VALUES (5, NULL, 20, 20)
GO

-- dati
SELECT * FROM dbo.position

/*

position_id location_id PLN_ID      fixel_id
----------- ----------- ----------- -----------
1           NULL        10          10
1           NULL        20          20
1           NULL        30          30
2           NULL        10          10
2           NULL        20          20
4           NULL        10          10
4           NULL        20          20
4           NULL        30          30
5           NULL        10          10
5           NULL        20          20
202         NULL        10          10
202         NULL        20          20
202         NULL        30          30
202         NULL        40          40
202         NULL        50          50

*/

-- implementazione senza cursore
-- funzione di ranking ROW_NUMBER(), conta finche la chiave di PARTITION BY è uguale, ordinando per i campi definiti in ORDER BY.
;WITH Incrementante AS
(
  SELECT
      position_id
    , location_id = ROW_NUMBER() OVER(PARTITION BY position_id ORDER BY position_id, PLN_ID, fixel_id)
    , PLN_ID
    , fixel_id
  FROM
    dbo.position
  WHERE
    position_id = 202
)

-- aggiornamento dalla WITH (common table expression - CTE)
UPDATE P
SET location_id = I.location_id
FROM
  dbo.position      P
  JOIN Incrementante    I  ON P.position_id  = I.position_id
                AND P.fixel_id    = I.fixel_id 
                AND P.PLN_ID    = I.PLN_ID 
            
-- dati in sorgente    
SELECT * FROM dbo.position ORDER BY position_id DESC, location_id 

/*

position_id location_id PLN_ID      fixel_id
----------- ----------- ----------- -----------
202         1           10          10
202         2           20          20
202         3           30          30
202         4           40          40
202         5           50          50
5           NULL        10          10
5           NULL        20          20
4           NULL        10          10
4           NULL        20          20
4           NULL        30          30
2           NULL        10          10
2           NULL        20          20
1           NULL        10          10
1           NULL        20          20
1           NULL        30          30

*/

-- pulizia
DROP TABLE dbo.position
GO