Accès libre

Multi-level cache management of quantitative trading platform

Zou Lida
Zou Lida
, 
Hassan A. Alterazi
Hassan A. Alterazi
 et 
Mohammed Helmi Qeshta
Mohammed Helmi Qeshta
   | 10 nov. 2021
Applied Mathematics and Nonlinear Sciences's Cover Image
À propos de cet article
Article précédent
Article suivant
Citez
Publié en ligne: 10 nov. 2021
Pages: 249 - 260
Reçu: 02 mars 2021
Accepté: 22 juil. 2021
DOI: https://doi.org/10.2478/amns.2021.2.00045
Mots clés
© 2021 Zou Lida et al., published by Sciendo
This work is licensed under the Creative Commons Attribution 4.0 International License.

Fig. 1

User experience function.
User experience function.

Fig. 2

The framework of multi-level cache management.
The framework of multi-level cache management.

Fig. 3

Time-series data merging.
Time-series data merging.

Fig. 4

The effect of cache capacity on user experience.
The effect of cache capacity on user experience.

Fig. 5

The effect of caching device on user experience. LRU, least recently used; LFU, least frequently used.
The effect of caching device on user experience. LRU, least recently used; LFU, least frequently used.

Fig. 6

The performance of the merging algorithm. LRU, least recently used; LFU, least frequently used.
The performance of the merging algorithm. LRU, least recently used; LFU, least frequently used.

Fig. 7

The performance comparison of different users.
The performance comparison of different users.

Cached Data Merging Operation (CDMO)

Input: C, P. C is the set of cached data for a stock s and formalized as C = {ci | i ∈ [0, n)}, where i is integer. Assuming that pi is the value of ci, P is formalized as P = {pi | i ∈ [0, n)}, where i is integer.
Output: new value of C and P after merging.
1. O ← P
2. WHILE O ≠ Ø DO
3.   pk = max{pi | pi ∈ O}, O←O–{pk}
4.   Q←O
5.   WHILE Q ≠ Ø DO
6.     qj = max {qi | qi ∈ Q}, Q←Q–{qj}
7.     IF ck and cj overlap
8.       merging ck and cj and the new data is cl
9.       IF pi ≥ pk
10         C←C–{ck, cj}+{cl}
11         P←P–{pk, pj}+{pl}
12.         RETURN CDMO (C, P) // Iteratively call Algorithm 1
13.   END WHILE
14. END WHILE
15. RETURN C, P

Multi-level Cache Placement Algorithm (MCP)

Input: C, E, W, where W is the set of device capacities, i.e., W = {wi | ei ∈ E}
1. E ← sort(E) //sort devices in decreasing order of read-white rate.
2. FOREACH ei in E
3.   FOREACH cj in C
4.     Compute value xj of cj
5.     X ←X ∪ {xj} //put xj into set of data value
6.   END FOREACH
7.   get the largest xj in X, place cj into ei, until wi runs out; update C, i.e., remove the cached cj in C; X ← Ø
8. END FOREACH
eISSN:
2444-8656
Langue:
Anglais
Périodicité:
Volume Open
Sujets de la revue:
Life Sciences, other, Mathematics, Applied Mathematics, General Mathematics, Physics
RSS Feed de la revue