No, I don't think that was the aim at all.
For reasons best known to your lecturer, I think they want you to store the hexadecimal number 0aah in all the memory accessible using the ds data segment register as the segment register.
There are a number of things with which you seem to be having trouble.
(1) hexadecimal numbers. So that the assembler recognises them, it expects a 0 on the start to indicate that it's a number of some sort (rather than a few letters), and the h on the end indicates hexadecimal. Default is decimal, so 099 will be taken as 99 decimal, while 099h will be understood as hexadecimal, (9*16)+9. If you aren't happy with hexadecimal, base 16 counting, find the appendix in a good assembler textbook.
(2) Memory addressing. In non-protected mode all memory addresses have a segment and an offset. The segment is actually simply shifted left 4 bits (= multiplied by 16) and added to the offset to create a 20-bit address which corresponds to 1Mbyte of memory, the DOS-addressable space (of which only about 500K is for programs. The rest is operating system/screen/etc space.
The segment bit is what's held in cs, ds, es. The offset can be in various registers or just a number, so when you do
mov myvariable, ax
what you are doing is putting the contents of ax into memory location addressed by ds:constant, where constant is the offset the assembler is using for your variable 'myvariable'.
(3) Filling the whole data segment with 0aah. This means putting 0aah in every byte accessible by ds. To address lots of locations like this, you probably want to use an instruction like
mov ds:si, al
where the al register already contains the value 0aah. You need to loop this instruction for all possible values of si.
But, important but, you (probably) cannot do this in real life. Your lecturer has set you a question whose answer cannot be tested. And that is something I find deplorable. All programs should be tested, and to encourage a student to do otherwise is a bad act! This sort of exercise is to be frowned upon!
The reason you can't do it in real life is you have no way to know how much of the data segment is available for your data, and how much already contains useful things. In some cases the data segment might actually overlap with the code segment (e.g. in com applications they're the same). Or it might overlap with the stack. Either would cause catastrophic disaster if things were overwritten.
It is never a good idea to write to memory whose function you do not know.
Good luck!