Adaptive precision setting for cached approximate values

Abstract

Caching approximate values instead of exact values presents an opportunity for performance gains in exchange for decreased precision. To maximize the performance improvement, cached approximations must be of appropriate precision: approximations that are too precise easily become invalid, requiring frequent refreshing, while overly imprecise approximations are likely to be useless to applications, which must then bypass the cache. We present a parameterized algorithm for adjusting the precision of cached approximations adaptively to achieve the best performance as data values, precision requirements, or workload vary. We consider interval approximations to numeric values but our ideas can be extended to other kinds of data and approximations. Our algorithm strictly generalizes previous adaptive caching algorithms for exact copies: we can set parameters to require that all approximations be exact, in which case our algorithm dynamically chooses whether or not to cache each data value.We have implemented our algorithm and tested it on synthetic and real-world data. A number of experimental results are reported, showing the effectiveness of our algorithm at maximizing performance, and also showing that in the special case of exact caching our algorithm performs as well as previous algorithms. In cases where bounded imprecision is acceptable, our algorithm easily outperforms previous algorithms for exact caching.

Online Algorithms to Minimize Resource Reallocations and Network Communication

“…The problem of setting the precision of approximated cached values is defined in [OLW01] as follows. There are data sources S 1 , .…”

“…We add a hard requirement that the sum of the lengths of intervals assigned to the data sources be at most a parameter T ; in [OLW01], this was a soft requirement. The goal is to maintain valid intervals of total size T in a way that minimizes the overall network traffic, assuming that each report of an invalid interval and each reassignment of intervals takes one message.…”

“…We call our version of this problem Transmission-Minimizing Approximate Value Problem (TMAV). [OLW01] give an optimal solution to a related problem under the assumption that the sources' values arise from certain types of probability distributions. Similar problems were studied in [CYV04,cKA02], who present heuristic approaches and experimentally evaluate their performance.…”

“…On the other hand, our lower bounds apply to online algorithms that have complete knowledge about past requirements. The Transmission-Minimizing Approximate Value problem is a generalization of one defined in [OLW01], in which low-power sensors monitor real-time events in a distributed wireless network and report their results to a centralized cache. In order to minimize network traffic, the cache is allowed to maintain approximations to the values at the sensors, in…”

Multi-Dimensional Online Tracking

“…In our problem, we compare with the offline algorithm that is allowed to send in any function value and at any time, as long as the error bound ∆ is satisfied. Finally, it should be noted that similar problems have been studied in the database community [10,13,15,17]. However, all the techniques proposed there are based on heuristics with no theoretical guarantees.…”